The biology of Samson hippos with emphasis on the sportfishery in Western .

By Andrew Jay Rowland

This thesis is presented for the degree of Doctor of Philosophy at Murdoch University

2009

DECLARATION

I declare that the information contained in this thesis is the result of my own research unless otherwise cited.

……………………………………………………. Andrew Jay Rowland

2 Abstract This thesis had two overriding aims. The first was to describe the biology of Samson Fish Seriola hippos and therefore extend the knowledge and understanding of the Seriola. The second was to uses these data to develop strategies to better manage the and, if appropriate, develop catch-and-release protocols for the S. hippos sportfishery. Trends exhibited by marginal increment analysis in the opaque zones of sectioned S. hippos otoliths, together with an otolith of a recaptured calcein injected fish, demonstrated that these opaque zones represent annual features. Thus, as with some other members of the genus, the number of opaque zones in sectioned otoliths of S. hippos are appropriate for determining age and growth parameters of this . Seriola hippos displayed similar growth trajectories to other members of the genus. Early growth in S. hippos is rapid with this species reaching minimum legal length for retention (MML) of 600mm TL within the second year of life. After the first 5 years of life growth rates of each sex differ, with females growing faster and reaching a larger size at age than males. Thus, by 10, 15 and 20 years of age, the predicted fork lengths (and weights) for females were 1088 (17 kg), 1221 (24 kg) and 1311 mm (30 kg), respectively, compared with 1035 (15 kg), 1124 (19 kg) and 1167 mm (21 kg), respectively for males. Despite these differences, female and male S. hippos attained similar maximum age, i.e. 29 (1470 mm FL) and 28 years (1280 mm FL), respectively. The maximum age determined for S. hippos is greater than that recorded for any other Seriola spp. The largest female and male S. hippos recorded during this study were encountered during the tagging component and had fork lengths of 1600 mm and 1380 mm, respectively. Seriola hippos has a protracted spawning period, ca four months, which starts in late spring and continues through summer into early autumn during which time many individuals engage in large spawning aggregations on the lower west coast of Australia. The length at which 50 % of the females in the population reached maturity was 831 mm FL (888 mm TL) and approximately 4 years of age, whilst all females over 950 mm FL were mature. Whilst aggregated for spawning S. hippos ceases feeding, however, during the non-spawning period this species can best be described as an opportunistic carnivore which feeds on a variety of pelagic and demersal prey. This study has greatly increased our understanding of S. hippos movement on the west and south coasts of Australia and has documented, for the first time, the migratory behaviour of a carangid in these waters. Many S. hippos individuals undertake long distance migrations to join spawning aggregation sites near Rottnest Island. Individuals

3 tagged at these aggregation sites where recaptured throughout this species distribution along the south coast of Australia, some after travelling distances of over 2400 km. Many S. hippos individuals displayed strong temporal and spatial spawning ground fidelity as numerous fish released at the spawning aggregations were recaptured at the exact same spawning site at similar times in subsequent years. Tagging data suggest that on the completion of spawning S. hippos individuals return to a resident location and remain in that general vicinity over the winter months. This study has developed a hypothesis describing larval dispersal associated with the S. hippos spawning behaviour exhibited near Rottnest Island. It is proposed that variations in the prevailing currents, at this important spawning location throughout the protracted spawning period, leads to high intra and inter-annual variation in larval distribution and survival. The affect of this variation on the evolution of the spawning and migratory behaviour displayed by S. hippos is discussed. A recent increase in the popularity of S. hippos as a catch-and-release sportfish has led to concerns by some anglers about post release survival of this species, particularly due to the depth of capture. Short term mortality of S. hippos capture at the sportfishing sites was assessed by monitoring fish held within an enclosure near the site of capture for up to 31 hours post release. The total hooking mortality of S. hippos subjected to catch- and-release within the Rottnest Island sportfishery is approximately 8%. Most of this observed mortality is delayed and occurs sometime after release. Although best handling practises require ongoing development, the current level of mortality associated with this catch-and-release fishery is considered acceptable. Furthermore, this mortality is likely to have little effect on the S. hippos population due to the high abundance of this species and the fact that even the highest fishing effort yields a relatively low catch. Seriola hippos exhibits a typical teleost neuroendocrine stress response associated with catch-and-release. The physiological dysfunction associated with the stress of capture in this species does not appear to cause any post release mortality. Instead, most mortality was attributable to barotrauma, however, although mortality in S. hippos increases with capture depth, this species is much less susceptible to depth induced mortality than other commonly targeted species in which barotrauma has been observed. This study developed key handling protocols for fishers who catch-and-release S. hippos at the Rottnest Island aggregation sites. These protocols cover aspects of catch- and-release fishing such as hook , water depth, time at surface, release method and shark .

4 Almost all S. hippos observed during capture from deep water released large quantities of gas from the opercular region, particularly during the last 10 to 20 m before reaching the surface. This phenomenon has also been witnessed by divers and fishers to occur under natural conditions. Investigations into this release of gas revealed this physoclistous species to exhibit unique swim bladder characteristics. Seriola hippos possess a membranous tube that connects the posterior-dorsal surface of the swim bladder internally to a region under each externally. This connection, termed the swim bladder vent, allows the escape of expanding swim bladder gases on rapid ascent. The presence of the swim bladder vent provides an explanation as to why the incidence of external barotrauma symptoms in S. hippos captured from the deepwater was low. The ability to expel excess swim bladder gases during rapid ascent whilst retaining full swim bladder function is likely to offer this semi-pelagic species considerable advantages when hunting prey, avoiding predators and engaging in spawning activities. Preliminary estimates of total mortality indicated that S. hippos is not currently subjected to a high level of fishing pressure. However, managers must remain mindful of the fact that the size at which females reach sexual maturity, i.e. 888 mm TL, is greater than the current minimum legal length, i.e. 600 mm, and thus fishers are currently allowed to harvest sexually immature fish. Furthermore, the effectiveness of future conservation measures must consider the large scale migration and spawning strategy undertaken by this species in to ensure its protection. The collaborative research approach undertaken during this study demonstrated that a high level of community engagement produced a large amount of research interest, increased stakeholder satisfaction from project input, improved understanding of research outcomes, and increased research uptake, all of which has led to increased stewardship and conservation of the S. hippos fishery and resources in general. Indeed, projects of this nature would not be possible without this type of approach.

5 Acknowledgements

My greatest thanks must go to my supervisors Drs Howard Gill and Mike Mackie for their guidance, support and giving me the opportunity to undertake this research. I am very fortunate to have had two supervisors who took a great interest in this research. Thank you both for sharing your incredible knowledge and expertise, for guided research focus towards applicable outcomes and for you eagerness to assist in the field. Thanks Howard for your efforts spent catching and tagging these splendid yet formidable fish, and thanks Mike for also testing your stamina as well as expertly skippering the often mechanically challenged Snipe. Many thanks must also go to Paul Lewis for his tireless work during the Samson Science project and for his problem solving wizardry. Thank you also for compiling the commercial catch data for Samson Fish. I wish to express my utmost gratitude to Allan and Yvonne Bevan. This study was greatly improved with your assistance and support. Thanks Al for helping collect samples, sharing your vast knowledge of the ocean, for the many arm stretching days fishing followed by black cans and for your friendship. Your enthusiasm for all fish species, but Sambos in particular, inspired me greatly. My sincere thanks to Frank Prokop for your advice, encouragement and enormous support over a number of years, and for teaching me the true meaning of stakeholder engagement and research uptake. I would also like to thank Con Costa, Mark Pagano, Kane Moyle and the board members of Recfishwest for their support. Thanks must also go to Garry Lilley and Wally Parkin for sharing their passion for the marine environment and the many late night discussions on all things fish, you are true champions for the development and promotion of responsible attitudes towards recreational fishing. Thanks to all the members of Murdoch University’s Centre for Fish and Fisheries Research who assisted in various aspects of my research, in particular, Dr Dean Thorburn, Mike Taylor, Peter Coulson, Dr Alex Hesp, Prof. Norm Hall and Prof. Ian Potter. I am also grateful for the expertise of Gordon Thomson for helping with histological preparations and Dr Kate Bryant for expert advice on evolution. Thanks to Dr Rod Lenanton for discussions on the intricacies of the Leeuwin Current and to Dr Kate Hutson for sharing her passion for kingfish. Great thanks are due to Bill Sawynok of Info-fish Services for his ongoing advice and encouragement, particularly during the community extension components of this study. The expertise of David Hall from Hallprint was also appreciated for assistance on the many aspects of fish tagging. I would also like to thank Joff Weston and the crew of Blue Juice, Ricky Lim, Brendon McConnell, Jim Palatchie, Paul Mckeown, Rusty Ellis, Damien Lane, Damian Langridge, the members of the Naturaliste Gamefishing Club and the Volunteer Fisheries Liaison officers who assisted with this research. Thanks to the members of Australian National Sportfishing Association in WA who provided tag data and supported the Samson Science project, in particular Steven Gilders, John Stevens, Richard Howell and Neil Dawes. The short term post-release survival experiments would not have been possible without the vertical enclosure or ‘sock’ borrowed from the Queensland Department of Primary Industries and Fisheries particular thanks go to Dr Ian Brown, Dr Wayne Sumpton and Mark McLennan. Many thanks also to Theo Berden and Mark Baxter who so skilfully skippered the RV Naturaliste whilst working with the sock. I am indebted all the members Western Australian recreational fishing community who volunteered their and supported the Samson Science project so enthusiastically. Many aspects of this research would not have been possible without your support and much is testament to the quality assistance that you supplied. Financial contribution for this research was kindly provided by the Australian Government’s Fisheries Research and Development Corporation, Murdoch University, the Department of Fisheries , and Recfishwest. Thank you to my Mum and Dad for their loving support and patience throughout all of my studies. Finally, I would like to thank my beautiful partner Fe for her wonderful love and companionship throughout this time.

Table of Contents

Abstract ...... 3 Acknowledgements ...... 6 Table of Contents ...... 7 Chapter 1 : Introduction ...... 10 1.1 General characteristics of ...... 10 Systematics ...... 11 Commercial and recreational importance ...... 12 Carangids in Australia ...... 12

1.2 The genus Seriola ...... 14

1.3 Seriola hippos in Australia ...... 17

1.4 The recreational and commercial fishery for Seriola hippos ...... 19 Recreational Fishery ...... 19 Rottnest Island Sportfishery ...... 21 Commercial Fishery ...... 24

1.5 Aims of the study ...... 27

Chapter 2 : Age, growth, mortality and reproductive biology of Samson fish, Seriola hippos Günther 1817, in Western Australia...... 30 2.1 Introduction ...... 30

2.2 Materials and Methods ...... 33 Collection of samples ...... 33 Initial measurements ...... 34 Otolith preparation and age determination ...... 35 Analysis of growth ...... 39 Reproductive variables ...... 40 Length at maturity ...... 42 Dietary data ...... 43 Mortality ...... 44

2.3 Results ...... 46 Description of Seriola hippos sagittae ...... 46 Validation of otolith increment periodicity - marginal increment analysis ...... 48 Validation of otolith increment periodicity – calcein injection ...... 49 Analysis of the precision between counts and the readability of otoliths ...... 50 Age and growth analyses ...... 51 Length – weight relationships ...... 58 Length frequency distributions ...... 58

7 Reproduction ...... 59 Sex ratios ...... 60 Spawning season and gonad development ...... 60 Length and age at first maturity ...... 65 Fecundity ...... 66 Mortality ...... 67 Diet ...... 71

2.4 Discussion ...... 73 Age and Growth ...... 73 Juvenile Ecology ...... 78 Reproduction and Spawning ...... 77 Diet ...... 80 Mortality ...... 81 Management ...... 82

Chapter 3 : Movement and migration of Samson fish, Seriola hippos Günther 1817, in Western Australia ...... 84 3.1 Introduction ...... 84

3.2 Materials and Methods ...... 88 Samson Science ...... 88 Tagging ...... 90 Leeuwin Current ...... 94

3.3 Results ...... 95

3.4 Discussion ...... 103 Migration of Seriola hippos ...... 104 Significance of Seriola hippos spawning migrations ...... 108 Oceanic processes of the lower west coast of Western Australia ...... 111 Rottnest Island spawning sites and localised oceanic processes ...... 112 Proposed larval distribution...... 114 Possible factors driving the reproductive strategy of Seriola hippos ...... 118 Adult migration and ocean currents ...... 121 Conclusions and future research ...... 122

Chapter 4 : Post release survival of Samson Fish, Seriola hippos Günther 1817, after catch-and-release angling...... 124 4.1 Introduction ...... 124

4.2 Materials and Methods ...... 131 Tag data ...... 131

8 Enclosure trials ...... 131 Blood physiology ...... 134

4.3 Results ...... 135 Tagging and engagement of fishers ...... 135 Enclosure trials – general observations ...... 137 Enclosure trials – survival ...... 138 Physical effects of capture ...... 141 Blood physiology ...... 143

4.4 Discussion ...... 145 Mortality ...... 145 Physiological stress response in Seriola hippos ...... 147 Handling Protocols ...... 149 Angler uptake and feedback ...... 158 Implications for research and management of recreational fisheries...... 159

Chapter 5 : A novel swim bladder allows rapid ascent in kingfishes (Seriola spp.).161 5.1 Introduction ...... 161

5.2 Materials and Methods ...... 167

5.3 Results ...... 168

5.4 Discussion ...... 172

Chapter 6 : Summary and General Conclusions ...... 176 6.1 The biology of Seriola hippos ...... 176

6.2 Movement and migration of Seriola hippos ...... 178

6.3 Post release survival of Seriola hippos ...... 181

6.4 The unique swim bladder of Seriola hippos ...... 184

6.5 Management implications ...... 186

References ...... 188

Appendix ...... 209

9 Chapter 1

General Introduction

1.1 General characteristics of Carangidae

The family Carangidae is a diverse group of which constitute species commonly known as jacks, scads, pompanos, queenfishes and trevallies, among others.

Carangids are actively swimming fishes, generally categorized as either neritic carnivorous bottom feeders or neritic planktivores, with most members moving in schools displaying semi-pelagic habits (Gunn 1990). A few species in this family, such as the

Rainbow Runner Elagatis bipinnulata and Pilotfish Naucrates ductor are pelagic and found in the open ocean often near the surface (Smith-Vaniz 1999). The body of carangids, although generally laterally compressed, is extremely variable in shape ranging from elongate and fusiform to deep and strongly laterally compressed. Juveniles and adults possess two separate dorsal fins (Figure 1.1). The anterior contains between 3 and 8 spines (becoming embedded in adults of some species), whilst the second dorsal fin has a long base and contains one spine and 18 to 44 soft rays. Carangids are distinguished from all similar families by the presence of two anterior pterygiophores that are elongate and bear spines which are separated from the rest of the anal fin by a gap (one spine in Elagatis and Seriolina), but which often become embedded in adults (Smith-

Vaniz 1999). Most carangids only possess small cycloid scales, which in many species are often modified into enlarged spiny scutes along the posterior section of the .

Members of this family possess a deeply forked caudal fin with equal upper and lower lobes and often have a slender caudal peduncle. These features provide highly efficient tail propulsion which generates the swimming mode known as carangiform, a characteristic named after the family. This mode of swimming involves considerable movement of the tail with the majority of thrust produced by lateral undulations confined to the posterior 30% of the body while the anterior bends only slightly (Webb 1975).

10

Separate Lateral line arched dorsal fin s anteriorly Second dorsal fin with long base

Caudal fin fork with lobe s of equal size Slender caud al peduncle of ten with scutes

Two spines separated from anal fin by gap

Figure 1.1. Characteristic used to diagnose members of the family Carangidae. Drawing from Smith-Vaniz 1999.

Systematics

The family Carangidae belongs to the suborder Percoidei in the order and is made up of about 32 genera incorporating 140 species worldwide (Nelson 2006).

Taxonomic relationships within the family are somewhat unclear and a considerable amount of nomenclatural change in the past has occurred as a result (Laroche et al. 1984,

Gunn 1990, Honebrink 2000). In a major review of the Carangidae, Smith-Vaniz (1984) produced a cladogram that identified four tribes based on morphological characters, the

Trachinotini, Scromberoidini, Naucratini and Carangini. Nelson (2006) provisionally recognised these tribes as subfamilies. Trachinotinae contains 21 species in the two genera Lichia and Trachinotus. Scromberoidinae is made up of the three genera

Oligoplites, Parona and Scomberoides, containing 10 species. Naucratinae is characterised by the presence of caudal peduncle groves dorsally and ventrally and includes the five genera Compogramma, Elagatis, Naucrates, Seriola and Seriolina, with

11 a total of 13 species. is the largest subfamily and is distinguished by having scutes present posteriorly along the lateral line and caudal peduncle (Smith-Vaniz 1984).

This subfamily comprises about 96 species in 22 genera, including Atule, ,

Caranx, , , and . Reed et al. (2002), using DNA sequences from the mitochondrial cytochrome b gene, found support for the monophyly of three subfamilies within the Carangidae (Caranginae, Naucratinae, and Trachinotinae), however these authors concluded that monophyly of the forth tribe (Scromberoidini) was questionable. Many carangid genera are currently being revised by W. F. Smith-Vaniz

(Paxton et al. 1989, Honebrink 2000) which will greatly aid in the clarifying the taxonomic relationships within this family.

Commercial and recreational importance

Worldwide, Carangidae are one of the most important families of commercial fishes taken by trawls, purse-seines, traps and long-lines, and all species are used for food

(Smith-Vaniz 1999). The Chilean Jack Trachurus murphyi is within the top ten fish species exploited worldwide and is now deemed fully exploited and over exploited in parts of the Southeast Pacific with a total catch of 2 508 834 metric tonnes in 2001 declining to 1 663 542 t in 2005 (FAO 2007). Larger species of carangids, particularly those of the genera (trevallies), Scomberoides (queenfish) Seriola (kingfish and amberjacks) and Trachinotus (pompanos) are highly regarded as sportfish (Cusack and

Roennfeldt 1988, Grant 1999, Smith-Vaniz 1999).

Carangids in Australia

Sixty three species of Carangidae from 23 genera occur in Australian waters, eight of which are endemic (Gunn 1990). Australia’s tropical regions hold the greatest carangid diversity with 52 species having distributions primarily north of 23°S (Figure 1.2) whilst only 6 species have temperate distributions, namely Samson Fish Seriola hippos,

12 Yellowtail Kingfish Seriola lalandi, Silver Trevally Pseudocaranx dentex, Sand Trevally

Pseudocaranx wrighti, Common Jack Mackerel Trachurus declivis and

Trachurus novaezelandiae (Gunn 1990, Gomon et al. 1994). The latter two of these species form a large part of Australia’s Small Pelagic Fishery (SPF) (previously known as the Commonwealth Jack Mackerel Fishery), where they are taken by purse seining and mid water trawling along with Blue Mackerel Scomber australasicus and Redbait

Emmelichthys nitidus. This Commonwealth co-managed fishery extends south from the

Queensland/New South Wales border to north of (31° S) with the most productive region being in eastern and southern Tasmania where, in 1986-87, the catch of T. declivis peaked at almost 40 000 t (Kailola et al. 1993). Historically, most SPF catches have been

Trachurus species which were processed into fishmeal, with some used as rock lobster bait. In more recent seasons, E. nitidus, comprising more than 70% of the catch, has replaced T. declivis as the main species caught in the SPF, of which much is now used as feed in Southern Bluefin Tuna Thunnus maccoyii aquaculture in South Australia (Findlay

2007).

Darwin

Northern N Territory Queensland apricorn Tropic of C 23 S Western Shark Bay Australia South Brisbane Moreton Bay Australia New South PACIFIC Wales Perth Great Australian OCEAN Rottnest Sydney Bight Adelaide Island Canberra Spencer Gulf Victoria Jervis SOUTHERN Melbourne Cape Leeuwin OCEAN Bay Yorke Peninsula 500 km Tasmania Hobart

Figure 1.2. Map of Australia showing the states and features referred to throughout this thesis. N.B. the tropical regions north of 23°S hold the greatest carangid diversity with 52 species, whilst 6 species have primarily temperate distributions (Gunn 1990).

13 1.2 The genus Seriola

Nine species of Seriola occur worldwide (Table 1.1). The three largest species of the genus, S. dumerili, S. hippos and S. lalandi, are reported to each reach lengths of over

180 cm TL and weights of 70 to 80 kg (Grant 1991, Allen 1997, Smith-Vaniz 1999). All species in the genus are epibenthic and pelagic and, whilst some tend to be solitary, most generally occur in small to medium sized schools. Small juveniles of each species have been recorded to be associated with floating structures, such as plants and debris, in oceanic and offshore neritic waters (Leis 1991, Castro et al. 2002, Nelson 2006). Three of the larger members of the genus, S. dumerili, S. lalandi and S. quinqueradiata, have been subjected to numerous studies due to their economic importance. Each of these species shows rapid growth rates, particularly in the first few years of life, and all are spring-summer spawners (Baxter 1960, Manooch and Potts 1997a, b, Gillanders et al.

1999a, b). Furthermore, each has been reported to migrate, with migration pattens stimulated by sexual maturity and water temperature (Baxter 1960, Kimura et al. 1994,

Thompson et al. 1999).

In Australian waters the genus is represented by four species, namely, S. dumerili,

S. hippos, S. lalandi, and S. rivoliana. Whilst the globally distributed, S. dumerili and S. lalandi have been subjected to numerous detailed studies, e.g. reproductive biology

(Marino et al. 1995, Gillanders et al. 1999b, Michale et al. 1999, Poortenaar et al. 2001,

Moran et al. 2007), food and feeding habits (Schmitt and Strand 1982, Matallanas et al.

1995, Sanderson et al. 1996), age and growth (Manooch and Potts 1997a, b, Gillanders et al. 1999a, Thompson et al. 1999, Kozul et al. 2001, Wells and Rooker 2004), movement and migration (Baxter 1960, Gillanders et al. 2001, Hutson et al. 2007a) and aquaculture

(Mazzola et al. 2000, Mylonas et al. 2004, Chen et al. 2007, Hutson et al. 2007b), there is a paucity of comparable data on the other two species that inhabit Australian waters.

14 Table 1.1. Species names, common names, maximum sizes and distribution of species in the genus Seriola. References: 1. Pizzicori et al. 2000, 2. Laroche et al. 1984, 3. Thompson et al. 1999, 4. Smith-Vaniz 1999, 5. Kozul et al. 2001, 6. Jover et al. 1999, 7. Nelson et al. 2004, 8. Smith-Vaniz 2002, 9. Corsini et al. 2006, 10. Kailola et al. 1993, 11. Grant 1999,12. Nakada 2002, 13. Love et al. 2005, 14. Sakakura and Tsukamoto 1999, 15. Lin and Shao 1997, 16. Okata 1976.

Species Common Names Maximum Size Distribution TL (cm) Kg S. carpenteri Guinean amberjack1 72.51 Sub-tropical waters in eastern Atlantic coastal regions and the Mediterranean Sea.1, 2 S dumerili. Greater amberjack3 1884 80.64 Circumglobal in sub tropical coastal waters and the Atlantic, Pacific Mediterranean amberjack5 and Indian .3 Mediterranean yellowtail6 S. fasciata Lesser amberjack7 67.58 4.68 Waters in the western Atlantic Ocean from Massachusetts into the Gulf of Mexico, also found in Cuba and Bermuda.8 Also found in the sup-tropical waters of the eastern Atlantic and the Mediterranean Sea.9 S. hippos Samson fish10 1734 7011 Temperate coastal waters of southern Australia and northern New Sea Kingfish10 Zealand.11 S. lalandi Yellowtail Kingfish1 19010 7010 Circumglobal with disjunct populations restricted to sub-tropical Yellowtail amberjack4 waters.4 Gold striped amberjack12 Yellowtail Jack13 Southern Yellowtail10 S. peruana Fortune jack7 Eastern Pacific Ocean.14 S. quinqueradiata Japanese amberjack15 Endemic to the north west Pacific Ocean in the coastal waters of Yellowtail16 and northern Hawii15 Amberfish16 S. rivoliana Almaco jack9 80 8 244 Circumglobal in tropical oceanic waters, entering sub-tropical Pacific amberjack9 waters in some areas. Oceanic, rarely in shallow water.4 Highfin amberjack11 S. zonata Banded rudder fish8 80 8 Found only in coastal waters over the continental shelf in the western Atlantic from Maine, U.S.A., to Santos, Brazil.8

1 Not only are Seriola spp. important commercial and recreational fish, with a worldwide catch in excess of 58 000 t in 1996, but they also support large aquaculture industries, particularly in Japan (FAO 1998). Seriola spp. are ideally suited to aquaculture as they display high growth performance, are hardy and have a highly regarded flesh, which in Japan is generally consumed as sushi or sashimi (Poortenaar et al. 2001). The main species within the Japanese industry, S. quinqueradiata, cultured there since 1927, currently relies on the capture of wild juveniles for on-growing in sea cages (Nakada

2002). Aquaculture production of this species in Japan has declined since its peak in

1995, when cultured S. quinqueradiata (170 000 t) was almost 3 times greater than the wild harvest (60 000 t) (Nakada 2002). This decline is due to a decrease in profitability of

S. quinqueradiata farms as over production has lead to a decrease in the market price at the same time that production costs have increased. Furthermore, Nakada (2002) highlights a large decline in the consumption of this species by young people in Japan.

Consequently there has been a shift recently in Japan towards the culture of S. dumerili and S. lalandi as these species have become more popular and desirable than S. quinqueradiata and thus obtain a higher price.

The culture of S. lalandi has recently been commercially established in Australia and , where it is reliant on hatchery reared juveniles (Poortenaar et al. 2003).

Production of this species in Australia is in its infancy and is largely confined to South

Australia where, in 2002, the South Australian Marine Finfish Farmers Association reported production of about 1200 t with a value of around $13 million (Love and

Langenkamp 2003).

Seriola lalandi also forms the basis of a large commercial fishery in New South

Wales. Landings in this fishery peaked in the late 1980s and early 1990s at between 450 to 600 t per year when the main fishery techniques were surface and subsurface traps, which accounted for about 80% of the catch (Kailola et al. 1993). However, catches of

16 this species dramatically declined when kingfish traps were banned in 1996 with recent catches of 145 t and 100 t reported in 2003-04 and 2004-05, respectively (ABARE 2006).

1.3 Seriola hippos in Australia

Samson Fish S. hippos also colloquially named sea kingfish, kingie, and sambo is a temperate, epibenthic species endemic to the southern half of Australia and the northern coastal waters of New Zealand (Gomon et al. 1994, Huchins and Swainston 2002, Prokop

2006). Seriola hippos is a fast swimming predator of the or open sea. It has an elongate, laterally compressed body with a forked caudal fin. The species varies considerably in coloration and body shape with growth (Figure 1.3). Small juveniles have a very blunt head and 5 broad black vertical bars on a yellowish to green background

(Figure 1.3a), a colouration which is likely to imitate floating algae around which juveniles of this species are found (Castro et al. 2002). These characters become less pronounced with age with adults displaying a less convex upper profile and more uniform colour (Smith-Vaniz 1999). Adult colouration is variable with blue green to dull purple- brown above, paler across the flanks with a sliver to white when live (Grant 1999)

(Figure 1.3c). This colouration generally darkens after death changing to brown-bronze with a pale belly. When viewed underwater the colour is usually silver to silvery grey and a dark stripe is often present extending from the posterior margin of the jaw through the eye to the origin of the first dorsal fin (Hutchins and Thompson 1995, A. Rowland pers. obs). In Australia, it has a discontinuous distribution that extends from Moreton Bay in southern Queensland (27° 25’S) to Jervis Bay, New South Wales (37° 39’S) on the east coast and from Yorke Peninsula (ca 35°S, 137°E), on the south coast, west, to Shark Bay in Western Australia (25° 21’S) (Paxton et al. 1989, Huchins and Swainston 2002) (Figure

1.2). This species is especially common along the coast of Western Australia where it

17 inhabits inshore oceanic waters and is often associated with reefs, wrecks and pylons, smaller fish are sometimes found in (Smith Vaniz 1999).

a)

b)

c)

Figure 1.3. Photographs of Seriola hippos a) a 49mm TL juvenile captured under floating Sargassum west of Rottnest Island, b) a 282mm TL juvenile caught near the mouth of the Swan River, Western Australia, c) a 910 mm TL adult caught in 40m depth off Lancelin, Western Australia.

18 In recent times S. hippos has become regarded by anglers as an excellent sportfish because it attains a large size (maximum total length of 173cm and a weight of up to 53 kg) and displays strong fighting abilities when hooked. In the “Sandgroper’s Guide To

Better Fishing”, Bodeker and Mulgrave (1980), state that “for sheer brute strength and dogged stamina, few southern fish can match the mighty samson fish”, explaining that

“The stab of a hook sends it straight for the horizon, a manoeuver that has stripped many a reel and unnerved many an angler”. These attributes, which are implied in the common name Samson Fish, are also reflected in the species name, hippos, meaning horse.

1.4 The recreational and commercial fishery for Seriola hippos

Recreational Fishery

Due to its poor reputation as a food fish, S. hippos is rarely targeted by recreational anglers for its eating qualities. Nevertheless, this species is common in recreational catches on the west and south coasts of WA due to its abundance and broad distribution.

Often this is as a by-catch when targeting prized demersal species such as Pink Snapper

Pagrus auratus and West Australian Dhufish hebraicum or when trolling lures or floating baits for pelagic species such as Spanish Mackerel Scomberomorus commerson or other scombrids such as tuna. Many boat fishers regard S. hippos as somewhat of a nuisance species as it has a habit of moving into a fishing area in numbers, taking baits, breaking lines and making it impossible to catch demersal target fish. This species is commonly targeted, however, during competitive line and spearfishing events for its high point scoring value because of its large size, and by catch-and-release sportsfishers.

There are little data on catches of S. hippos by recreational fishers. Surveys conducted in the West Coast Bioregion (WCB) (Figure 1.4) during 1996-97 (Sumner and

Williamson 1999) indicated that 5687 (66%) S. hippos individuals were kept by boat

19 anglers whilst 2934 (34%) were released. The National Recreational and Indigenous

Fishing Survey undertaken in 2000/01 (Henry and Lyle 2003) estimated an annual harvest of 10,890 fish in WA for a catch described as Kingfish/Samson Fish, with an estimated weight of 98 642 kg. It is highly likely that the majority of these fish were S. hippos as opposed to the less common S. lalandi or S. dumerili which were also included in this grouping (Kailola et al. 1993).

* * * * * *

Figure 1.4. Map indicating Bioregions used in the management of Western Australian fisheries resources as defined by the Department of Fisheries W.A.. Squares are Catch and Effort System (CAES) reporting blocks for the commercial fishing sector. Areas within the West Coast Bioregion used for assessment of the commercial catch of Seriola hippos are shown in the inset. Asterisks denote the CAES block data used in Figure 1.7. 20 Rottnest Island Sportfishery

Seriola hippos is recognised for its game fish attributes and targeted by sportfishers within the West Coast Bioregion (Figure 1.5), particularly in waters of around 100 m depth approximately 10 to 15 km west of Rottnest Island where the species aggregates during summer (November to March). These aggregations, which form over shipwrecks in what is known as the Rottnest Island Graveyard, are the basis of what has recently developed into a world renowned sportfishery (Bain 2007). The Rottnest Island

Graveyard which spans an area of approximately 25 km2 is located in water depths between 60 and 120m. Many vessels that were deemed to be no longer of use have been scuttled in this area since 1910, and then after the second world war surplus military equipment, submarines and commercial aircraft were also dumped there (Green 2005). It appears that this material has provided artificial habitat for many fish species and, within this area, there are at least 14 such sites where large numbers of S. hippos gather during the summer months (A. Bevan pers comm.). The locations of only a small number of the sites, however, are known to most fishers.

Prior to the 1990s this species was rarely targeted except by members of the local gamefishing community who, in the mid 1980s and early 1990s, would specifically fish for this species for catch-and-release. These anglers normally fished in an area on the southern side of Rottnest Island called Kingfish or Kingie reef, where they would anchor and attract S. hippos with a berley/chum of fish pieces (Cusack and Roennfeldt 1988).

Whilst the Rottnest Island aggregation sites were also occasionally fished during this time, the target species was not S. hippos. Instead gamefishers would stop to fish these locations for demersal species such as Pink Snapper on their return trip to port after a day spent targeting tuna Thunnus spp. and marlin Makaira spp. west of Rottnest Island.

During the mid 1990s two charter boats based in Fremantle began to specifically target S. hippos for the purpose of catch-and-release sportfishing (A. Bevan pers. comm., Shikari

21 Charters). Clients for these day trips targeting these fish were mainly international visitors from Asian countries such as Japan and Malaysia, where the deep water jigging style of fishing for similar species is popular.

Figure 1.5. Photographs of Seriola hippos caught at the aggregation sites west of Rottnest Island. Fish of this size and condition are typical of the sportfishery targeting deep water aggregation sites. Note the variation in colour.

Over the following years these charter boat operators would also stop at the S. hippos aggregations for short periods with local clients to allow standard demersal charter groups the experience of catching these powerful fish. In 1998 one operator stopped targeting S. hippos due to concerns over the mortality of released fish (see below). It was at this time the charter operators commenced discussion with scientists at the Department of Fisheries Western Australia (DoFWA) that ultimately led to the current research.

The remaining charter operator continued targeting S. hippos west of Rottnest

Island, building an international reputation for specialist catch-and-release sportsfishing charters. Tagging of fish became an integral component of these charters as a means to aid research. During this period the aggregation sites were visited on irregular occasions by larger private boats, typically chasing this species during fishing competitions. By

2001 two other charter boats started making intermittent trips to the aggregations, as the popularity of catch-and-release fishing increased among international visitors and local fishers.

22 Recognition of this fishery steadily increased through local fishing media exposure and promotion by tackle stores and by 2003 private boats were regularly undertaking trips to specifically target S. hippos west of Rottnest Island for catch-and-release. A fishery that was once only within reach of knowledgeable skippers with commercial grade, and thus expensive, electronics needed to find fish in deep water is now a popular fishery with local fishers as information about the fishing methods and aggregation locations has become common knowledge. However, access to the Rottnest Island S. hippos aggregations is generally limited to boats greater than 6 m in length due to the large distance offshore and the strong sea breezes commonly encountered in this area during summer. Interest in this fishery has also increased due to the depleted state of traditionally targeted demersal species and the associated decline of quality fishing experiences.

Most anglers targeting these large fish have adopted the methods promoted through the fishing media, based on rod and line fishing with deep water jigs - the technique also used by most charter operators. This method uses a painted metal or lead jig rigged with a single barbless hook attached to a heavy braided line of 24 to 40 kilograms breaking strain. Jigs are usually attached to the braided line via a heavy monofilament leader (70 to 90 kg breaking strain), which provides shock absorption when a fish strikes the jig and also allows the fish to be effectively controlled when near the boat for timely release. Jigs weighing between 200 and 500 grams rigged by this method are dropped vertically from the vessel to the bottom and are retrieved in a jerking motion to imitate prey. Schooling S. hippos are very receptive to this method of fishing and are readily caught. Although uncommon, this species is also captured using a baited rig with hook and sinker at sportfishing sites.

Anglers expressed concern for the survival of released S. hippos within the sportfishery, as mentioned earlier, due to the potential for barotrauma because of the depth of water in which the aggregations are found (> 100m). Barotrauma is the physical effects

23 of rapid and extensive reduction in environmental barometric pressure (Philip 1974, St

John 2003). The associated syndrome in humans is referred to as the bends or decompression sickness. Barotrauma in fish refers to body tissue and organ injuries caused by a sudden decrease in ambient pressure resulting from rapid accent to the surface during capture from deep water. Benthic dwelling physocylists fish, where the swim bladder does not directly connect to the digestive tract, are especially susceptible to barotrauma. In particular, rapid decompression in a large number of species causes swim bladder overexpansion and potential damage to associated organs, exopthalmia (bulging eyes), internal and external haemorrhaging, everted stomach, intestinal protrusions, and loss of equilibrium (Bruesewitz et al. 1993, Parrish and Moffitt 1993, Wilson and Burns

1996, Rummer and Bennett 2005). External symptoms of barotrauma have been recorded in fish captured in water as shallow as 3.5m (Shasteen and Sheehan 1997). Although the incidence of external barotrauma symptoms in S. hippos captured from the deep water aggregation sites is low, it is very common for angled fish to release large amounts of gas when nearing the surface. It was this phenomenon that prompted initial angler concern as the internal effects of decompression appeared significant but remained uncertain because most S. hippos individuals seemed to swim away strongly when released.

Commercial Fishery

Seriola hippos is not considered as a major commercial target species in Western

Australia. Commercial catches, however, increased rapidly during the early 1980s reaching a peak of just over 126 tonnes in 1987 (Figure 1.6). Most of the catch during this time was taken in the Mid-West and South Coast regions (Figure 1.3), and most was taken by hand and dropline. In 2005 the total commercial harvest of S. hippos in Western

Australia was 86.6 t (mean 1995-2005 = 102.1 t), with 33.2 and 26.4 t of this captured from the Mid-west and Metropolitan regions, respectively (DoFWA unpublished data)

24 Figures 1.3). Although not a species of major commercial interest, S. hippos does contribute an important portion to the West Coast Demersal Scalefish Fishery. For instance, in 2004/05, S. hippos (50 t) ranked fifth in the top species caught within that fishery behind P. auratus (333 t), Lethrinus spp. (199 t), G. hebraicum (196 t) and

Centroberyx spp. (83 t) (St John et al. 2005). The annual commercial harvest of S. hippos is similar to that estimated for the recreational sector. For instance, assuming an average fish weight of 10 kg, the total commercial catch in Western Australia for 2006 would have consisted only 8660 individuals.

140000

120000

100000

80000

60000

40000 Catch Live Weight (kg) Weight Live Catch

20000

0

5 1 7 8 9 9 1 1977 1979 1 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Handlining Droplining Longlining YearNet Trawling Charter Other

Figure 1.6. Annual commercial catch of Seriola hippos by fishing method in Western Australia from 1977 to 2005.

Most of the S. hippos catch in the Metropolitan region is likely to be taken from aggregations that occur in deep water west of Rottnest Island (M. Mackie, pers. comm.,

DoFWA), however, it is worth noting that generally there is little overlap in commercial and recreational fishing activities for this species. Although peaks in commercial catches

25 occur in spring-summer when S. hippos form large aggregations along the west coast, they are also caught throughout the year in most areas (Figure 1.7).

60

50

40

30 Catch (t) 20

10

0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month

31150 Nth Rotto 32150 Sth Rotto 33140 Capes 30140 Jurien 29140 Dongara 28130 Sth Abrol

Figure 1.7. Total commercial catch of Seriola hippos (1979-2006) per month for the main CAES blocks (indicated in Figure 1.4). Boxed area indicates time of the year when aggregations of this species occur on the west coast.

The value of landed S. hippos varies markedly depending on the quality of fish and demand. The price at market for whole fish ranges from $0.80/kg for gillnet bycatch with lice damage to $3.50/kg for premium smaller line caught fish. There is generally a low demand for S. hippos and it is not uncommon for catches to be passed in when offered at the Perth Fish Markets, when sold however the typical market price is $1.50 to $2.00/kg for whole fish (P. McKeowen pers. comm., Kailis Bros.). The target market for S. hippos landed commercially is mostly fish processors who fillet and sell onto supermarkets where it is retailed as low-priced fillets generally labelled as kingfish. Premium quality fish are also purchased by processors or seafood retailers to be sold as fillets or fish for sashimi.

26 1.5 Aims of the study

This thesis has two overriding aims. One offers a scientific overview of the species, whilst the second involves practical applications towards the fishery.

1) Describe the biology of S. hippos and therefore extend the knowledge and understanding of the genus.

2) Investigate the effects of catch-and-release fishing on S. hippos targeted at the

Rottnest Island sportfishing sites and, if appropriate, develop catch-and-release protocols for this species.

To fulfil these overriding aims the age, growth and reproductive biology of S. hippos were described and the following hypotheses tested.

• It is hypothesised that S. hippos will display similar growth trajectories to other

Seriola spp. which have been demonstrated to attain large size and display similar

semi-pelagic habits. Therefore, the first objective of this thesis was to provide key

aspects on the age and growth of S. hippos (Chapter 2).

• The second objective of this thesis was to, in describing its reproductive biology,

test the hypothesis that S. hippos form the deep water aggregations west of

Rottnest Island for spawning purposes (Chapter 2).

• The temporal nature of the large aggregations near Rottnest Island together with

evidence of long distance movement from early tagging of S. hippos captured near

Rottnest Island suggests that many S. hippos individuals move long distances to

join the annual deep water aggregations, i.e. they are transient spawning

aggregations sensu Domeier and Colin (1997). Furthermore, other large Seriola

27 spp. have been documented to undertake long distance movements seasonally

(Baxter 1960). Thus, the present study aimed to test this hypothesis by extending

tagging efforts to investigate the movement and any possible migration of S.

hippos associated with the large, deep water aggregations west of Rottnest Island

(Chapter 3).

• Barotrauma symptoms associated with line capture in deep water are rarely seen in

S. hippos. Furthermore, anecdotal evidence from sportfishers targeting deep water

S. hippos aggregations suggest this species is robust and individuals generally

swim away strongly when released. In addition, S. hippos that had been tagged at

the aggregations prior to this study and recaptured up to several years later

demonstrates that fish can survive capture from deep water. It is therefore

proposed that S. hippos show a high rate of survival after this type of catch-and-

release angling. Thus, the present study aimed to test this hypothesis by

investigating the effects of catch-and-release fishing on S. hippos at the

aggregations sites west of Rottnest Island. If this hypothesis is accepted the data

generated will then be used to develop catch care and fish handling protocols to

maximise the survival of released fish (Chapter 4).

• Although anglers considered S. hippos to be a robust fish that displayed low

incidences of barotrauma and released well after capture, they also reported that

gas bubbles were frequently released by fish as they were brought to the surface.

Anglers, divers and commercial lobster fishermen also reported this phenomenon

in free swimming fish undertaking rapid ascent to the surface. As carangids

possess a physoclistous swim bladder, S. hippos would be expected to suffer

barotrauma during these rapid ascents. The release of gas on rapid ascent may be a

result of swim bladder rupture or S. hippos possessing some

28 morphological/physiological adaptation that negates the effects of rapid decompression. These competing hypotheses are tested in Chapter 5.

29 Chapter 2

Age, growth, mortality and reproductive biology of Samson Fish, Seriola hippos Günther 1817, in Western Australia.

2.1 Introduction

Reliable biological information on age structure, growth, mortality rates, length and age at first maturity, spawning period and fecundity is essential for developing fisheries management plans for a species. Numerous studies have documented such parameters in three of the larger members of the Seriola genus. For example, aspects of the reproduction and growth of the circum-globally distributed Seriola dumerili were reported for populations from the Gulf of Mexico by Manooch and Potts

(1997a), Thompson et al. (1999) and Wells and Rooker (2004), from the western North

Atlantic Ocean by Manooch and Potts (1997b) and Harris (2004), from the Adriatic Sea by Kozul et al. (2001) and from the South Mediterranean Sea by Marino et al. (1995).

Similarly, the reproductive biology and growth of Yellowtail Kingfish Seriola lalandi have been documented from Californian waters (Baxter 1960), from the waters of New

South Wales on Australia’s eastern coast (Gillanders et al. 1999a, b, Stewart et al. 2004) and from the east and west coasts of New Zealand (Poortenaar 2001, Walsh et al. 2003,

30 Moran et al. 2007). The Japanese Amberjack Seriola quinqueradiata, a species endemic to the waters of Japan and northern (Lin and Shao 1999), has, in particular, been the subject of many biological studies (e.g. Nakada 2002 and references therein) because of its aquaculture significance. Despite such focus on the genus, there are no published data on any aspect of the life history of the Samson Fish Seriola hippos.

Age determination studies within the genus have used a variety of calcified structures, such as scales, vertebrae, opercular bones, dorsal spines and otoliths (Baxter

1960, Gillanders et al., 1999a, Thompson et al. 1999, Kozul et al. 2001, Manooch and

Potts 1997a). Certain species of Seriola, e.g. S. dumerili and S. lalandi, are notably difficult to age (Gillanders et al. 1999a, Manooch and Potts 1997a) which is reflected by the numerous structures and procedures undertaken by various researchers.

Gillanders et al. (1999a), in an ageing study focused on various calcified structures in S. lalandi, concluded that, otoliths can be used for reliably age determination as long as validation of annuli formation on a regular time scale is possible. Interestingly, Gillanders et al. (1999a) found that growth zones on whole otoliths were better for ageing S. lalandi than zones on sectioned otoliths as the latter could be rarely interpreted. This finding is rather unusual as many studies have found that it is necessary to section otoliths in order to reveal all of the growth zones (Beamish 1979). This is because whole otoliths display asymmetric deposition of material which, when read, can underestimate ages due to difficulties in detecting outer growth zones (Beamish and McFarlane 1987). Gillanders et al. (1999a), however, only aged a small number of fish greater than 90 cm fork length (n

= 20) with a maximum age of 9 years and suggested that more reliable age estimates of larger fish may be attained using sectioned otoliths. In a later study involving a greater number of larger fish, Stewart et al. (2004) successfully used sectioned otoliths to describe size-at-age for S. lalandi in commercial landings and observed a maximum age of 21 years. Sectioned sagittal otoliths have also been used on numerous occasions to

31 successfully determine the age of S. dumerili (Manooch and Potts 1997a, b, Thompson et al. 1999, Harris 2004).

Although there have been few detailed studies, all Seriola species examined to date are reported to have a spring-summer spawning period (Baxter 1960, Marino et al. 1995,

Manooch and Potts 1997a, Gillanders et al. 1999a, b, Michale et al. 1999, Harris 2004).

Knowledge of the spawning period enables the determination of reliable growth parameters as it allows a realistic birth date to be assigned to fish of all ages.

Trends shown by gonadosomatic indices (GSIs) throughout the year is one of the most common methods used to describe the reproductive cycles of fish as it can provide an indication of gonadal maturation and the particular spawning strategy of a species

(West 1990). However, GSI trends do not provide a precise indication of when spawning actually occurs and need to be considered with other measures of gonadal development.

Macroscopic gonad staging supported with microscopic staging of histologically prepared sections, which, although time consuming and expensive, allows for measurement and staging of oocytes, are proven techniques in reliably assessing stages of gonadal maturation (West 1990, Gill et al. 1996, Mackie and Lewis 2001). When used in combination, these methods can provide efficient and cost effective data on the spawning period and spawning frequency as well as the age and size at sexual maturity of a species.

The length at which a species first reaches maturity is often used to set the minimum legal length (MLL) at which that particular species can be legally retained if caught (Hill 1990). Minimum legal lengths, combined with gear restrictions, are central elements of fisheries management strategies focussed at both commercial and recreational fisheries (Winstanley 1990). In Western Australia S. hippos is grouped with S. lalandi and

S. dumerili for management purposes and all species have been allocated a MLL of 600 mm total length by the Department of Fisheries Western Australia. Furthermore, current regulations restrict recreational fishers to retaining a maximum of two Seriola species

32 combined per day in all waters off Western Australia (Anon. 2008a, b, c and d). As the size at maturity of S. hippos is not known, it is not possible to determine whether or not the current MLL is appropriate.

In light of its commercial and recreational importance in Western Australia, which was highlighted in Chapter 1, the overall aim of this chapter was to provide, for the first time, details on critical aspects of the biology of S. hippos. The individual aims were as follows:

1. Evaluate sectioned sagittal otoliths for ageing this species by validation of annual opaque zone (annuli) formation.

2. Determine fish lengths at specific ages and derive growth parameters.

3. Estimate instantaneous rates of total, natural and fishing mortality.

4. Determine the duration and peak time of spawning and describe ovary and oocyte development during this period.

5. Determine age and length at which maturity is attained.

6. Determine if fecundity is determinate or indeterminate and provide fecundity estimates.

The results of this chapter will be considered in the context of the development of appropriate management plans for this species in Western Australia. This chapter also aims to test the hypotheses that 1) S. hippos will display similar growth trajectories to other Seriola species and, 2) S. hippos is a summer spawner and that the large aggregations found west of Rottnest Island are indeed spawning aggregations.

2.2 Materials and Methods

Collection of samples

Fresh samples of S. hippos were collected on regular intervals from sites near Perth

(31°57’S, 115°51’E) onboard recreational, charter and research vessels from coastal waters up to 200m in depth between January 2004 and February 2007. Additional

33 samples of S. hippos were collected from commercial fish processors and at recreational fishing competitions throughout this period and frozen frames (fillets removed) were collected from recreational fishers. Seriola hippos were also collected opportunistically from Geraldton (28°47’S, 114°37’E), Albany (35°01’S, 117°53’E), Bremer Bay (34°23’S,

119°23’E) and Esperance (33°52’S, 121°53’E) (Figure 1.3). Sampling for small juvenile

S. hippos (252 to 600mm FL) was conducted in inshore waters of the Perth region where fish were collected from shallow reefs, channel markers and jetty pylons. Juvenile fish, or fish of undifferentiated sex (26 to 65mm FL), were collected on occasion from floating weed mats (predominately Sargassum spp.) using a fine mesh scoop net in waters west of

Rottnest Island from 50 to 200m in depth and from the near shore waters of Perth.

Initial measurements

The total length (TL) and fork length (FL) of each S. hippos was measured to the nearest 1 mm. In addition to the standard measurements the head length (HL, from tip of snout to firm edge of operculum) and jaw length (JL, from tip of snout to posterior edge of upper jaw) of each fish were also measured (Figure 2.1). When whole fish were sampled the whole weights of fish < 10 kg were weighed to the nearest 1 g, while those > 10 kg were weighed to the nearest 50 g. The weights of fish that could not be obtained prior to filleting were estimated from the regression equation that relates wet weight to fork length

(see Results). The relationship between HL and JL with FL was calculated (see Results) so that the FL of an individual could be estimated when only the head of the fish was available, as was sometimes the case for large fish obtained from recreational fishers.

34 Total length

Fork length

Head length

Jaw length

Figure 2.1. Details of length measurements made on Seriola hippos. The position of cut used to remove sagittal otoliths is also indicted.

Otolith preparation and age determination

The two sagittal otoliths of S. hippos were removed by cutting vertically into the dorsal side of the skull slightly posterior to the brain with a bone saw (Figure 2.1). This action cut through the otic capsules and the fragile otoliths were then carefully removed with forceps. After removal, otoliths were cleaned, dried and stored in Eppendorf vials within labelled paper envelopes. Eppendorf vials were perforated to allow otoliths to dry completely during storage.

Examination of whole otoliths of a small number (n=6) of small juvenile S. hippos

(41 to 65mm FL) using transmitted light on a compound microscope revealed clearly discernable growth zones. Although some growth zones were visible in the whole otoliths of larger S. hippos (i.e. >250mm FL), these were often hard to detect and the total number was not discernable. Growth zones in the otoliths of these fish became far more visible after sectioning which greatly enhanced their readability, thus, the otoliths of all S. hippos

>250mm FL were sectioned prior to counting growth zones.

A single otolith from each fish was selected and mounted in clear epoxy resin.

Prior to sectioning the core region of each otolith was identified under a dissecting

35 microscope and then marked on the surface of the resin. A single 500-600μm thick section containing the core was then cut using an Isomet Buehler low-speed saw with a diamond wafering blade. The section was then ground with fine wet-and-dry paper (1200 grade), washed, dried and mounted on a microscope slide using DePX mounting medium.

These sections were then examined with transmitted light and reflected light using a compound microscope. The number of opaque zones on each sectioned otolith was counted on two occasions without any knowledge of the fork length, specimen ID or date of capture of the fish from which the otolith had been removed. A section was examined for a third time when the two counts of the opaque zones differed. When the third count agreed with one of the previous two counts it was used as the age estimate. When all three readings differed the sample was rejected. It should be noted that an opaque zone was counted as an annuli even if it had not completely formed at the margin of the otolith (i.e. there was no translucent zone distally). During examination each otolith section was also assigned a readability index category (Table 2.1).

Table 2.1. Readability index (RI) categories used for Seriola hippos (after Buckworth 1999).

RI Category Annuli Readability 1 Unreadable 2 Poor 3 Fair 4 Good 5 Perfectly Readable

Validation of the assumption that recognisable annuli form annually in the sagittae of S. hippos was tested using several techniques. Validation of annulus formation using marginal increment analysis was undertaken, however, the commonly employed technique which involves measuring the distance between the outer edge of the outermost opaque zone and the edge of the otolith as outlined by Campana (2001) proved to be difficult due to sometimes indistinct translucent zones and often broad opaque annuli. Likewise, in an ageing study of S. dumerili, marginal increment analysis proved futile for Thompson et al.

36 (1999) due to an inability to determine the condition of the otolith edge. Otoliths in this study were therefore assigned a marginal increment category (Table 2.2), in a similar method to that used by Gillanders et al. (1999a) for S. lalandi.

Table 2.2. Otolith marginal increment (MI) categories used for Seriola hippos (after Lewis and Mackie 2002).

MI Category Otolith section margin appearance 1 Opaque zone at margin 2 1-50% of previous translucent zone 3 50-100% of previous translucent zone

In addition to marginal increment analysis, validation of annuli periodicity within otoliths was attempted by chemical marking of otoliths. Seriola hippos (n = 146) captured from near Rottnest Island were injected with calcein solution (25mg ml-1) and tagged with a specifically labelled nylon headed dart tag (Hallprint PDA, 120mm long, 3mm diameter) informing anglers to retain recaptured fish. Calcein is a fluorochrome dye which binds to calcium and fluoresces under ultraviolet light. The calcein solution was injected with a syringe and a 22 gauge needle into the coelomic cavity in close proximity to the region at a minimum dose of 15 mg kg-1 body weight. Fish weight was estimated from a table developed from the regression equation that relates wet weight to fork length (see

Results).

Tag and recapture data obtained during the present study (see Chapter 3) were also used to verify age estimates. Fish that had been at liberty for more that 300 days and showed positive growth (n = 119) were used to predict length at recapture from growth curve parameters (Labelle et al. 1993). The growth model for mark-recapture data developed by Baker et al. (1991) that incorporates size at tagging and time at liberty, and which is analogous to the Schnute (1981) size-at-age model, was used to predict length at recapture:

/1 b ⎧ −− tta mr )( ⎫ b −− tta mr )( bb a −− ττ 12 )( 1− e = ⎨ mr []−+ 12 eLLeLL a −− ττ )( ⎬ ⎩ 1− e 12 ⎭

37 where Lm and Lr are sizes at tm, age at marking, and tr, age at recapture, respectively. The parameters a, b, L1, L2, τ1 and τ2 are the Schnute (1981) age-length growth model parameters estimated from otolith annuli counts (see below).

Predicted lengths were then compared graphically to observed recapture lengths.

Although Francis (1995) provides an alternative, yet mathematically equivalent, analogue of Schnute’s growth model to that of Baker et al. (1991) which is considered preferable for analysing fish growth from mark-recapture data, Francis (1995) states that the analogue of Baker et al. (1991) is superior for estimating expected growth as it is simpler and, thus, more convenient. It must also be noted that this procedure is not statistically rigorous, as growth parameters are determined from age-length data where variation is assumed in length as a function of age, thus growth models are not used to predict age (the independent variable) from length (the dependent variable) (see Francis 1988a, b).

However, this method is useful for comparative purposes and should reveal any obvious departures from fitting growth curves based on the assumption that observed growth zones in the otoliths are deposited annually (Labelle et al. 1993). In addition to this method, the actual growth of eight individual specimens measured at both tagging and recapture was compared to the growth curve determined from the lengths at age for males and females combined. Fork length measurements at time of tagging were used to estimate age at tagging of individuals from growth curve parameters. Time at liberty and length at recapture were used to determine growth of tagged individuals. The growth of these recaptured fish was then compared graphically with the growth curve determined from the lengths at age using otolith annuli counts.

The birth date assigned to S. hippos was determined from the approximate time of peak spawning as estimated from mean monthly gonadosomatic indices and gonadal maturity stages. The birth date was used in conjunction with the otolith annuli count, marginal increment category and date of capture to allocate an absolute age to each fish.

38

Analysis of growth

Growth curves were fitted to the individual lengths of each female and male at their estimated age of capture. The lengths at age of juvenile fish that could not be sexed were allocated randomly, but equally, to the female and male data sets used for calculating the growth curves. The growth of S. hippos was initially analysed using the von

Bertalanffy growth equation, as is normally employed for describing the growth of most teleosts. The von Bertalanffy growth equation is:

-k(t-to) Lt = L∞ (1-e ),

Where Lt is the length at age t, L∞ is the mean asymptotic length, k is the growth coefficient and t0 is the hypothetical age at which fish would have zero length.

Although the von Bertalanffy growth curves of each sex fitted most of the length at age data well, they passed below the points for many larger fish. The Schnute growth equation was therefore fitted to the same length at age data as the von Bertalanffy growth equation to ascertain whether it provided a significantly better fit. The Schnute growth equation is:

/1 b ta −− τ1 )( ⎡ bbb 1− e ⎤ t ⎢ −+= LLLL 121 )( a −− ττ )( ⎥ ⎣ 1− e 12 ⎦

(Schnute 1981), where L1 and L2 are the estimated lengths at selected reference ages τ1 and

τ2 (years) and a and b are constants (both ≠ 0). From this equation the asymptotic length can be calculated using the equation:

/1 b aτ 2 b aτ1 b ⎡ 2 − LeLe 1 ⎤ L∞ = ⎢ aa ττ ⎥ ⎣ 2 − ee 1 ⎦

(Schnute 1981). 39 The data for each equation were fitted by minimising the sum of squared deviation between observed and predicted lengths using Solver in MicrosoftTM Excel.

The von Bertalanffy and Schnute growth curves derived for female and male S. hippos were compared using a likelihood-ratio test to determine whether there were significant differences between the curves derived for each sex using the different growth equations and between the curves derived for both sexes using the same growth equation

(Kimura 1980). The hypothesis that the data in each case could be described by a common growth curve was rejected at the α = 0.05 level of significance if the test statistic, calculated as twice the difference between the log-likelihood obtained by fitting a common growth curve for both sexes and by fitting separate growth curves for each sex, exceeded

2 χα (q), where q is the difference between the number of parameters in the two approaches.

(e.g. Cerrato 1990).

Reproductive variables

In the field, fresh gonads of S. hippos were removed and macroscopically staged soon after capture. On the basis of macroscopic characteristics each gonad was allocated to one of the following stages of gonadal development according to the staging system used by Mackie and Lewis (2001), for Spanish Mackerel Scomberomorus commerson and other species:

Undifferentiated: J (juvenile). Females: Stage 1 (virgin/immature), Stage 2-3 (mature resting), Stage 4 (reproductively developed), Stage 5 (spawning). Males: Stage 1 (virgin),

Stage 2 (mature resting), Stage 3 (reproductively developed/ripe), Stage 4 (spawning).

Where possible gonads were weighed to the nearest 1 g and the whole gonad or a 5 cm thick section from the mid region of one lobe was preserved in 10% neutrally buffered formalin solution and retained for histological examination. In the laboratory gonads were

40 rinsed and dried, embedded in paraffin wax, cut transversely into 6μm sections and stained with Mallory’s trichrome.

The microscopic staging system of Mackie and Lewis (2001) was used in the analysis of the preserved, histologically prepared samples. Although this system has more stages than the macroscopic system, it is compatible, and allowed for a more detailed description of reproductive development and spawning:

Females: Stage 1 (virgin/immature), Stage 2 (mature resting), Stage 3 (mature developing), Stage 4 (reproductively developed), Stage 5a (pre-spawning), Stage 5b

(spawning/running ripe), Stage 5c (post spawning), Stage 6 (spent).

Females were used for analysis since ovarian developmental stages are easier to distinguish than those of the testes and because ovarian development generally defines the spawning season (West 1990).

The gonadosomatic index (GSI) of individual S. hippos were calculated from the equation W1/(W2-W1) x 100, where W1 = wet weight of the gonad and W2 = wet weight of the whole fish, i.e. W2-W1 = somatic plus gut weight. The GSI was calculated using data for female fish > the estimated L50 at first maturity (see below). Where whole weight could not be measured an estimate of whole weight from the length-weight equation was used. It is acknowledged that GSI values determined using these weight estimates are approximate only.

Histological section of the ovaries of three mature female (stage 4 gonads) S. hippos individuals were analysed to assess the spawning mode of this species. The maximum and minimum diameters of 100 randomly selected oocytes, in which the nucleus was visible, were measure to the nearest 1 μm. The mean diameter of each oocyte was then calculated. The resultant oocyte diameter distributions, in conjunction with the stages of development of the oocytes, were used to determine whether S. hippos has determinate or indeterminate fecundity.

41 Batch fecundity is an estimate of the potential number of eggs released during one spawning event and is not an estimate of the total number spawned throughout the spawning season. Batch fecundity was estimated from oocyte counts in ovarian tissue that had been determined to be pre-spawning (stage 5a) through histological techniques. Both ovaries where removed from formalin solution, rinsed, then each was slit to allow excess fluid to drain, and the outer membrane dried with paper towel before being weighed to the nearest 1 mg. A tissue sample of between 250 and 300 mg was then taken from the middle, anterior and posterior regions of one ovary and weighed to the nearest 1 mg.

These tissue samples contained part of the outer membrane plus the connected ovarian tissue since membrane weight is also included in the whole gonad weight. Each sample was placed on a glass slide and covered with several drops of glycerol and examined under a dissecting microscope (x 10). Oocytes were loosened by gently teasing apart the tissue with forceps, the sample spread over the slide and the number of hydrated oocytes recorded. Hydrated oocytes were easily distinguishable from other non-hydrated oocytes by their large size and wrinkled appearance when preserved in formalin. The total number of hydrated oocytes (i.e. batch fecundity) for each female was calculated from the number of hydrated oocytes in each of the three pieces of ovarian tissue of known weight in conjunction with the total ovary weight.

Length at maturity

Females captured during the spawning period with gonads possessing stage 3 through to stage 6 oocytes were classified as mature because they had the potential to , were spawning or had spawned. The lengths at which 50 and 95% of females were mature or had begun to show signs of maturation (L50 and L95) were determined by fitting a logistic equation to the percentage contribution made, during the spawning period, by immature and mature S. hippos in each 50 mm length increment. These data were

42 randomly resampled and analysed to create 1000 sets of bootstrap estimates for the parameters of the logistic regression analysis and estimates of the probability of maturity within the range of recorded lengths. The point estimates of the L50 and L95, and of each probability of maturity at the specified length, were taken as the medians of these 1000 bootstrap estimates, and the 95% confidence limit of each parameter was calculated as the

2.5 and 97.5 percentiles of those 1000 estimated values.

The logistic equation used to predict the length at which females commenced

1 maturation was: P = , []−−+ 50 − LLLL 5095 )/())(19ln(exp1 where P = the proportion mature, L = fork length in mm, L50 and L95 are lengths in mm at which 50 and 95% of female fish reach sexual maturity, respectively, and ln is the natural logarithm.

Dietary data

The stomachs of S. hippos were removed in the field and either frozen or preserved in 100% ethanol. Stomach contents were later examined in the laboratory with the aid of a dissecting microscope. Diets of S. hippos were analysed using the frequency of occurrence and points methods (Hynes 1950, Ball 1961). The frequency of occurrence

(%F) method represents the frequency with which a particular prey type is consumed by a species, whilst the points methods gives the relative contribution of each prey type to the volume of the stomach contents of the fish. Stomach fullness was estimated on a scale of zero to 10, with zero representing an empty stomach, eight representing a full stomach and

10 representing a fully distended stomach. Each prey item was identified to the lowest possible taxon. The mean percentage volumetric contribution (%V) of each of these prey categories to the stomach contents was then calculated. The graphical methods of Costello

(1990) were used to display the diets of S. hippos, whereby percentage occurrence (%F) is plotted against relative quantity, in this case mean percentage volume (%V). Although

43 many dietary analyses use only contribution data, such as volume or weight measures, which are considered reliable data for dietary compositions (Hyslop 1980), the procedure of Costello (1990) also takes into account percentage occurrence and therefore gives insight into population wide food habits (Cortés 1997). This visual approach is used to indicate the importance of dietary categories by their position on a graph, with the more important categories by either %F or %V being separated from the minor categories.

Despite the fact that this method is simple to perform, it provides a basic assessment of dietary diversity and feeding trends for single species analysis and in some cases provides more robust conclusions than other more complicated analyses, such as the weight resultant index (Marshall and Elliott 1997). Only data from fish collected from within the west coast region were used for analysis as many fish collected from the south coast region had been gutted at sea or had the stomachs removed during the filleting process.

Mortality

The instantaneous rate of total mortality, Z, was determined for S. hippos using catch curve analysis (Hall et al. 2004). This analysis assumes constant annual recruitment and constant mortality after the age at which fish were considered fully vulnerable to fishing (Hall et al. 2004). The value for Z was estimated by maximising the log-likelihood using the SOLVER routine in Microsoft TM Excel. The data for S. hippos were randomly resampled, with replacement, and analysed to create 1000 sets of bootstrap estimates. The point estimate for Z was taken as the median of these 1000 bootstrap estimates, and the

95% confidence limit was calculated as the 2.5 and 97.5 percentiles of those 1000 estimated values.

Estimates for the instantaneous rate of natural mortality, M, were calculated from the relationship between natural mortality, growth and water temperature as described by

Pauly (1980). A regression of the same form used by Pauly (1980) was fitted to Pauly’s

44 data for 175 fish stocks using Microsoft TM Excel and inserting the values of k (years) and

L∞ (cm FL) and water temperature (°C) for S. hippos in temperate WA. The regression was fitted in the Statistical Package for the Social Sciences (SPSS inc., Chicago) to obtain a point estimate and associated 95% confidence limits for M. A mean annual surface water temperature for Perth coastal waters of 19.5oC was used for calculations determining mortality (Pearce et al. 1999).

Natural mortality was also estimated using relationships derived from Hoenig

(1983). Because the Hoenig equation for fish uses the relationship between total mortality and maximum age, this estimate of mortality equates to M if fishing pressure is light. As anecdotal evidence suggests this is the case for S. hippos, it was considered an appropriate estimate of M for this species in WA waters. A regression of the same form used by Pauly

(1980) was refitted to the data provide for 82 stocks described by Hoenig (1982). The maximum recorded age was then inserted into SPSS to obtain a point estimate and associated 95% confidence limits, thereby taking into account the variation in the data around the regression line and the uncertainty in parameter estimates. These two estimates of M were then combined using the likelihood distribution approach of Hall et al. (2004).

The Bayesian method described by Hall et al. (2004) was then used to determine the likelihood for M, calculated using the likelihood of Z, as determined from the catch curve analysis. This procedure assumes that, estimates of M are take from a uniform distribution in which M

A Monte Carlo resampling approach was used to determine the instantaneous coefficient of fishing mortality, F. Estimates of Z (derived from catch curve analysis) and

M (derived from the Bayesian approach) were randomly resampled, with replacement, from their respective probability distributions, but were discarded when corresponding values for M were greater than for Z. These values were used to produce 5000 estimates

45 for F, determined using the equation F = Z – M. The point estimate for F and associated

95% confidence limits were taken as the median value and the 2.5 and 97.5 percentiles of the 5000 estimates derived from this analysis.

2.3 Results

Between January 2004 and February 2007 a total of 714 S. hippos from three regions of Western Australia were collected and processed for biological analyses (Figure

2.2). Fork length (FL) of fish collected ranged from 26 (1.2 g) to 1470 mm (47 kg). The majority of these fish were collected from aggregation sites west of Rottnest Island (n =

317) and inshore waters in the Perth metropolitan area (n = 168). Other samples were sourced from throughout this species’ range in WA, including 89 fish from the Geraldton region in the north and 129 fish from southern coastal waters. Small juvenile fish (<100 mm FL, n = 22) were commonly associated with floating debris and jellyfish in offshore

(> 20 km from the coast, 56%) and inshore (44%) waters.

Description of Seriola hippos sagittae

The sagittae of S. hippos are small and fragile. They have a long narrow pointed rostrum with a shorter pointed anti-rostrum and a straight ventral margin with a rounded posterior (Figure 2.3a). The otolith is laterally compressed with a convex distal surface and the edges are generally smooth or have fine growth projections ventrally. A deep, prominent sulcus traverses the proximal surface longitudinally. A series of translucent and opaque zones are often distinguishable on the posterior section of the distal surface of whole dried otoliths when viewed under reflected light. These were validated as being laid down annually (see below). Although the first and second annuli were usually discernible, the total number was not detectable in the whole otoliths of most fish when compared with sectioned otoliths.

46 16 Perth region 14 inshore 12 n = 168 10 8 6 4 2 0

16 Perth offshore 14 spawning region 12 n = 317 10 8 6 4 2 0

16 North

Frequency (%) 14 n = 89 12 10 8 6 4 2 0

16 South 14 n = 129 12 10 8 6 4 2 0 0 200 400 600 800 1000 1200 1400 Fork Length (mm)

Figure 2.2. Length frequency distributions for Seriola hippos collected for biological analysis from inshore metropolitan, offshore metropolitan, northern and southern regions of Western Australia. n = sample size.

47 When viewed with transmitted light in cross section the sagittae of most adult fish also revealed a distinct series of narrow opaque and wide translucent zones (Figure 2.3b).

The distance between the first two opaque zones was typically wider than the distance between subsequent opaque zones, however, the first annulus was generally less distinct and often hard to distinguish.

a)

b)

Figure 2.3 Sagittal otoliths of Seriola hippos a) a whole dried otolith with line indicating position of section, b) a transverse section viewed with transmitted light.

Examination of whole otoliths of six small juvenile S. hippos (41 to 65mm FL) using transmitted light on a compound microscope revealed 29 to 58 rings. Due to the small sample size these could not be validated as daily. However, such rings in S. dumerili have bee shown to be laid down on a daily basis (Wells and Rooker 2004).

Validation of otolith increment periodicity - marginal increment analysis

Marginal increment analysis of sectioned S. hippos otoliths (Figure 2.4) demonstrated that an opaque zone was formed between September and March, with otoliths in this category (opaque margin) most abundant between December and January.

Those with category 2 margins (1-50% of the previous translucent zone) were present

48 from January to October and were most abundant in April and May whilst category 3 (51-

100% of the previous translucent zone) were present from March to December and most abundant in September and October. Thus, the marginal increment categories demonstrate a logical progression throughout the year with each category peaking once for 3-5 months.

A single opaque zone is therefore formed annually in the otoliths of S. hippos during spring/summer which becomes surrounded by a translucent zone during autumn. As the spawning of S. hippos occurs in late spring/summer (see below) the first opaque zone formed on the otoliths of 0+ S. hippos occurs towards the end of the first year of life.

MI 1 (Opaque) 100 MI 2 (1-50% Translucent) MI 3 (51-100% Translucent)

80

60

40 Percentage

20

0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec (54) (52) (68) (8) (14) (16) (14) (13) (20) (18) (41) (51) Month

Figure 2.4. Monthly percentages of each otolith marginal increment category for Seriola hippos. Data pooled by region, sex and age-class with the number samples for each month given in parentheses.

Validation of otolith increment periodicity – calcein injection

One hundred and forty six S. hippos captured from the Rottnest Island aggregations were injected with calcein solution and tagged with specific tags informing anglers to retain recaptured fish. Although five of these fish were recaptured by recreational fishers, only one fish was recovered for analysis. This specimen was a 940

49 mm FL male tagged and injected on the 2nd February 2006 and recaptured during the following summer on the 17th December. The time at liberty was 319 days during which this fish had gained 40 mm in length. A calcein mark was observed on a sectioned sagittal otolith of this fish between the two most recent opaque zones (Figure 2.5). Delineation of a translucent zone can be seen on the sectioned otolith soon after the time of calcein deposition. This translucent zone is then followed by a partly formed opaque zone at the otolith margin. The information obtained by this recapture revealed that a single annulus is formed during summer, as confirmed by marginal increment analysis.

A

A A

A

Ca

Figure 2.5. Sectioned otolith of a tagged and recaptured Seriola hippos after injection of calcein observed under ultraviolet light. Fish was at liberty for 319 days. Ca = calcein mark, A = annuli.

Analysis of the precision between counts and the readability of otoliths

Fifty seven (12.7%) of the 450 sectioned otoliths were rejected as being unreadable at the completion of the first two reading (Table 2.3). Of the fish rejected for ageing

83.9% were less than 1000mm FL. The agreement between the initial two counts for the remaining 393 otoliths was 80.7%. Otoliths for which the age differed between readings were re-examined and, in all cases, except one, the third reading corresponded to one of

50 the previous readings. Although the otoliths of S. hippos were not easy to read (i.e. most readability classed as fair), with the males having lower readability than females, most otoliths (87.3%) contained discernible annuli (Table 2.3).

Table 2.3. Percentage of sectioned Seriola hippos otoliths within each readability category.

Readability Categories (percentages) Sex n 1 (unreadable) 2 (poor) 3 (fair) 4 (good) 5 (excellent) Female 229 10.9 12.7 53.3 22.3 0.9 Male 191 14.7 18.3 56.5 10.5 0 Unknown 30 13.3 16.1 53.3 16.7 0 Overall 450 12.7 15.3 54.7 16.9 0.4

Age and growth analyses

Trends exhibited by reproductive variables (see section on Reproduction) demonstrated that the spawning period for S. hippos occurred in late spring/early summer peaking in November and continuing into January. The approximate mid-point of this period, i.e. 1st December, was therefore assigned as the birth date of this species.

The age distributions of female and male S. hippos aged from sectioned otoliths overlapped considerably (Figure 2.6). The oldest fish aged was a female at 29 years, with the oldest male being 28 years, 97% of fish aged were less than 20 years old and 68% were less than 10 years old. Female and male S. hippos are almost equally represented in fish greater than 20 years old contributing 55% and 45%, respectively.

51 30 Females Males 25

20

15 Frequency 10

5

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Age Class (years)

Figure 2.6. Age distribution for male and female Seriola hippos determined from sectioned otoliths. nmales = 167, nfemales = 207.

The likelihood ratio test demonstrated that the Schnute growth curve significantly improved (P<0.001) the fit for the length at age data for female and male S. hippos compared to that obtained using the von Bertalanffy equation (Figure 2.7). The Schnute growth curve overcame deficiencies at the upper end of the age range for each sex (Figure

2.7a). The improvement of fit by using the Schnute curve is emphasised by the fact that the coefficient of determination increased from 0.906 to 0.929, and 0.930 to 0.943, for females and males, respectively (Tables 2.4 and 2.5). It also resulted in a marked increase in the L∞ for both females and males from 1279 to 1594 mm, and 1139 to 1210 mm, respectively. Thus, the Schnute growth equation also provides a more realistic estimate of asymptotic length as the largest female and male individuals aged in this study had fork lengths of 1470 mm and 1280 mm, respectively, whilst the largest female and male encountered during the tagging component had fork lengths of 1600 mm and 1380 mm, respectively.

52

a) Schnute 1600

1400

1200

1000

800

600 Fork Length(mm) 400

200

0 0 5 10 15 20 25 30 Age (years)

b) von Bertalanffy 1600

1400

1200

1000

800

600 Fork Length (mm) Fork Length 400

200

0 0 5 10 15 20 25 30 Age (years) Females Males Female growth curve Male growth curve

Figure 2.7. a) Schnute and b) von Bertalanffy growth curves fitted to the lengths at age for males (dotted line and open circles) and females (solid line and closed circles) of Seriola hippos. nmales = 167, nfemales = 207.

53

Table 2.4. Von Bertalanffy growth parameters for female and male Seriola hippos derived from lengths-at-age of individuals (L∞ asymptotic length; k growth coefficient; t0 hypothetical age at length 0; R2 coefficient of determination; n sample size).

-1 2 k (years ) L∞ (mm) t0 (years) R N Females 0.188 1279 - 0.900 0.906 207 Males 0.240 1139 - 0.563 0.930 167

Table 2.5. Schnute growth parameters for female and male Seriola hippos derived from lengths-at-age of individuals (L1 and L2 lengths at selected reference ages τ1 (1 year) and τ2 2 (10 years); a and b constants (both ≠ 0); L∞ asymptotic length; R coefficient of determination; n sample size).

2 L1 (mm) L2 (mm) a b L∞ (mm) R n Females 435.3 1089.1 0.044 2.748 1594 0.929 207 Males 400.3 1034.5 0.136 1.971 1210 0.943 167

The likelihood ratio test demonstrated that the Schnute growth curve for females differed significantly to that of males (p<0.001). The lengths derived for female and male

S. hippos, using the Schnute growth equation, demonstrate that the growth of both sexes was rapid during the first 5 years. However, even at this early age females are bigger, with estimated lengths at ages 2, 4 and 5 for females being 608, 800 and 867 mm, respectively, compared to 566, 767 and 834 mm for males. Growth slowed markedly in both sexes after this period, particularly in males as growth trajectories became even more divergent. Thus, by 10, 15 and 20 years of age, the predicted lengths for females were

1088, 1221 and 1311 mm, respectively, compared with 1035, 1124 and 1167 mm, respectively, for males (Figure 2.7a). The maximum fork length and age recorded for female S. hippos was 1400 mm and 29 years, respectively, while for males it was 1280 mm and 28 years, respectively. Female and male S. hippos both attained the minimum legal length for retention after capture (MLL) of 600 mm TL (equivalent to 533 mm FL) within the second year of life.

54 Tagging conducted during the present study yielded 119 recaptures useful for comparison of growth rates. These data included a wide range of sizes (505 mm to 1400 mm FL) and times at liberty (302 – 1797 days). Observed lengths at recapture were generally close to those predicted and fell along the trajectory where predicted lengths equalled observed lengths (Figure 2.8). Variations in observed and predicted lengths were both positive and negative.

1600

1400

1200

1000

800

600

400 Observed fork length (mm) length fork Observed

200

0 0 200 400 600 800 1000 1200 1400 1600 Predicted fork length (mm)

Figure 2.8. Predicted versus observed recapture lengths of tagged Seriola hippos (n = 119). Predicted recapture lengths were based on Schnute growth parameters. The solid line covers points were the predicted and observed lengths are equal.

The growth of a sub-sample of eight recaptured S. hippos, were illustratively represented alongside the Schnute growth curve determined by the using otolith annuli counts (Figure 2.9). These fish, from a wide range of sizes (260 mm to 1193 mm FL at time of tagging), were at liberty for 403 to 1753 days. The growth trajectories created from the lengths at recapture and time at liberty for the eight individuals closely track the

Schnute growth curve. Thus, the tag recapture data support the age estimation technique and the lengths at age based growth model used during this study.

55 1600 Females

1400

1200

1000 Males 800

600 Fork (mm) Length

400

200

0 0 5 10 15 20 25 Age (years)

Figure 2.9. Growth trajectories of 8 tagged Seriola hippos individuals displayed with the Schnute growth curves fitted to the lengths at age for males, females and the sexes combines (solid line). Age at tagging was estimated from observed length at tagging using the Schnute equation and growth parameter determined from lengths at age data using otolith annuli counts.

Length – weight relationships

The likelihood-ratio test demonstrated that the relationship between caudal fork length

(FL, in mm) and whole weight (WW) for the two sexes were not significantly different

(p>0.05) (Figure 2.10). Therefore the data for males and females were combined giving a relationship between FL and WW of:

WW (g) = 1.497x10-4 x FL2.982 (mm) (n = 264, R2 = 0.99),

The relationship between TL and WW was developed for use within the catch-and- release sportfishery so anglers can estimate weight from fish length to reduce the amount of onboard handling before release:

WW (kg) = 4.98x10-9 x TL3.09 (mm) (n = 249, R2 = 0.92)

56 Note that only fish with a FL > 700 mm were used to develop this relationship and

associated conversion table (Table 2.6) as the inclusion of smaller fish tended to

underestimate the weight of fish typically caught in the sportfishery (i.e. 900 to 1300 mm

FL, see Figure 2.11a):

50000

45000 Females Males 40000

35000

30000

25000

20000

Whole Weight (g) 15000

10000

5000

0 0 200 400 600 800 1000 1200 1400 1600 Fork Length (mm)

Figure 2.10. Relationship between fork length and whole weight for male and female Seriola hippos. nmales = 103, nfemales = 161.

Table 2.6. Reference for the conversion of total length and fork length to weight for Seriola hippos of the size typically targeted by sportfishers.

Total Length Fork Length Whole Total Length Fork Length Whole (mm) (mm) Weight (kg) (mm) (mm) Weight (kg) 900 808 6.8 1300 1174 21.2 950 853 8.0 1350 1220 23.9 1000 899 9.4 1400 1266 26.7 1050 945 11.0 1450 1312 29.8 1100 991 12.7 1500 1358 33.0 1150 1037 14.5 1550 1404 36.6 1200 1083 16.6 1600 1450 40.3 1250 1129 18.8

57 The relationship between TL and FL for S. hippos was:

TL = 1.09FL + 17.84 (n = 443, R2 = 0.998)

Head length (HL) and jaw length (JL) to FL relationships were investigated to determine whether these could be used to provide FL estimators for the head only samples collected in this study. The relationship between HL and FL for S. hippos was:

FL (mm) = 4.34HL (mm) – 52.45 (n = 248, R2 = 0.981)

The relationship between JL and FL for S. hippos was:

FL (mm) = 10.22JL (mm) – 19.0 (n = 248, R2 = 0.957)

The HL based equation was used for FL conversion for the head only samples as there was less variance with this measurement. Head length is also generally easier to measure and is less influenced by individual variation in the augmentation and shape of the jaw.

Length frequency distributions

Lengths of S. hippos tagged by recreational anglers and researchers at the Rottnest

Island aggregations (n = 6886) during the summers of 2004/05 and 2005/06 (see Chapter

3) ranged from 550 to 1600 mm FL, with a median length of 1070 mm FL (Figure 2.11a).

The sex of 1762 of these fish was determined before release, and the sex of an additional

834 was determined at these locations during the summer of 06/07 (Chapter 3). Males (n

= 1150) ranged in length from 605 to 1380 mm FL while females (n = 1446) ranged from

780 to 1520 mm FL (Figure 2.11b). Female fish dominated the larger size categories

(Figure 2.11b) and had a median length of 1100 mm FL while males had a median length of 1040 mm FL.

58

a) 1400

1200

1000

800

600

400

200

0

Frequency 300 b)

250

200

Females 150 Males

100

50

0 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 Fork Length (mm)

Figure 2.11. Length frequency distributions of a) all Seriola hippos tagged and released (n = 6885) and b) Seriola hippos for which sex was determined prior to tag and release (nmales = 1149, nfemales = 1446) at Rottnest island aggregation sites. Black and grey dashed lines represent L50 and L95 at first maturity for females, respectively.

Reproduction

A total of 552 gonads from males, females and juveniles were collected and staged macroscopically during this study. Of these, 205 were processed for histological examination which demonstrated such macroscopic classification was valid (Figure 2.12).

59 The reproductive system of female and male S. hippos comprises two simple, elongate ovaries or testes that are joined posteriorly with a short gonoduct leading to the urogenital pore. During peak spawning the gonads of mature individuals of each sex occupy a majority of the body cavity volume. The weight of ovaries in the samples ranged from 1.2 to 3337 g, whereas testes ranged from 0.8 to 2240 g.

Sex ratios

Sex determination through cannulation of S. hippos captured at the aggregation sites near Rottnest Island (2004 – 2007) revealed a sex ratio of females to males of 1.26:1

(n = 2596). The sex ratio of female to male S. hippos collected for biological analysis in this study was 1.36:1 (n = 552).

Spawning season and gonad development

The gonadal stages in the corresponding months for the different regions were pooled for analysis due to the low number of gonad samples obtained. The mean monthly gonadosomatic indices (GSI) for females (> L50 at first maturity, i.e. 831mm) remained low (~1) from April to September. During October the indices rose sightly as ovaries developed. GSI reached a peak of ~6% in November and stayed at this level until

January, coinciding with a high proportion of ovaries being stage 5 (i.e. spawning) at this time (Figure 2.13). Even though many females were still spawning during February

(Figure 2.14) the large drop in GSI shows that the supplies of vitellogenic oocytes within the ovaries were reduced. This drop in GSI continued until May. These data confirm S. hippos individuals form the large aggregations observed west of Rottnest Island for spawning purposes as the timing of ovarian development coincides with the influx of large numbers of this species into the area from October to March.

60 Microscopic Characteristics Histological Section Stage Newly formed ovary F1 T Virgin/ containing very little ovarian tissue. Only Immature chromatin nucleolus stage oocytes (Cn) are present. As the ovary develops chromatin Cn nucleolus and preinucleolus stage oocytes increase in number and fill the Lu lamellae (La). Tunica La (T) thin. Lu – Lumen.

F2 Lamellae highly P Mature organised containing chromatin nucleolus resting (Cn) and perinucleolar stage oocytes (P – named because of the peripheral location of Cn the multiple nucleoli within the nucleus). This stage is seen soon after the preceding spent (F6) stage.

F3 Commences with the Developing development of cortical alveoli (yolk P vesicle) stage oocytes Cn (Ca). Chromatin nucleolus (Cn) and perinucleolar stage oocytes (P) are also Ca present. This stage Ca ends with the appearance of yolk

globule stage oocytes. 200 µm F4 This stage starts with the development of Developed Ca yolk globule stage oocytes (YG). YG Cortical alveoli stage oocytes (Ca) are also abundant. This is the 500 µm background state of ovaries during the spawning period.

Figure 2.12. Histological characteristics of the stages in the development of the ovaries of Seriola hippos following the microscopic staging system of Mackie and Lewis 2001. Stained with Mallory’s trichrome.

61 Microscopic Characteristics Histological Section Stage F5a Development into this Pre- stage is marked by the migration of the Hy spawning nucleus to the periphery of the oocyte , coalescence of yolk globules and the occurrence of hydrated oocytes (Hy). Ends with the ovulation of hydrated Hy eggs into the lumen. 500 µm

F5b A short and rarely observed stage when released eggs are found in the lumen and Spawning/ newly formed post ovulatory follicles (POFs) are found in the lamellae. Ovulated eggs are often not evident in the lumen after histological processing. Running Ripe

F5c This stage appears similar to the Post-spawn Ca developed (F4) stage but is defined by the presence of post ovulatory follicles (POFs). At ovulation YG the egg is expelled from these follicles into the ovarian POF lumen. YG - yolk globule stage oocytes, Ca - cortical alveoli 200 µm stage oocytes. F6 > 50% atresia of late yolk globule stage Spent Ca oocytes (YG) is the main criteria for this YG short stage between developed and resting stages. Lamellae appears disorganised containing chromatin nucleolus and perinucleolar oocytes and remnant cortical alveoli oocytes (Ca). 500 µm

Figure 2.12. Cont. Histological characteristics of the stages in the development of the ovaries of Seriola hippos following the microscopic staging system of Mackie and Lewis 2001. Stained with Mallory’s trichrome.

62 8

7 18 27 8 6

5

4 25

3 5 31 Gonadosomatic Index 2 5 6 10 12 13 1 4

0 Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Month

Figure 2.13. Mean monthly gonadosomatic indices ± 1 SE for females Seriola hippos > L50 at first maturity. Sample sizes in each month are given.

The ovaries of all female S. hippos > L50 at first maturity caught on the west and south coasts between April and August were stage 2 (Figure 2.14). Fish were first observed with developing (stage 3) or developed (stage 4) ovaries during early spring

(September), and by November 75% of fish encountered had developed gonads. Fish possessing spawning (stage 5) gonads were also first caught in November and were encountered in the ensuing months until March. The percentage frequency of individuals with stage 5 ovaries was highest in January (32%) and February (31%). Seriola hippos with spent and recovering ovaries (stage 6) were found between January and March

(Figure 2.14).

63 100 January 75 n = 38 50 25 0 100 February 75 n = 39 50 25 0 100 March 75 n = 42 50 25 0 100 April 75 n = 4 50 25 0 100 May 75 n = 10 50 25 0 100 June 75 n = 15 50 25 0 100 July 75 n = 18 50

Frequency (%) Frequency 25 0

100 August 75 n = 11 50 25 0 100 September 75 n = 15 50 25 0 100 October 75 n = 11 50 25 0 100 November 75 n = 20 50 25 0 100 December 75 n = 32 50 25 0 F2 F3 F4 F5 F6 Gonad Stage

Figure 2.14. Monthly percentage frequency of occurrence of sequential gonadal stages in female Seriola hippos > L50 as determined from histological sections. n = sample size in each month.

64 Length and age at first maturity

During the spawning season the ovaries of all S. hippos individuals with lengths

<700 mm FL were immature, i.e. stages 1 and 2 (Figure 2.15). Fish with ovaries at stages

3 to 6 were first recorded in the 700-750 mm FL length class in which they contributed

33%. All individuals in the 950-1000 mm FL and subsequent length classes were mature.

From the logistic regression analysis the length at first maturity (L50) for female S. hippos was 831 mm FL, which, on the basis of the Schnute growth equation, corresponds to an age of ca 4 years (Figure 2.15, Table 2.7).

385743355711911913127 42 5 100

75

50 Frequency (%) Frequency

25

0 400 500 600 700 800 900 1000 1100 1200 1300 1400 Fork Length (mm)

Figure 2.15. Frequency distribution of female Seriola hippos sampled during the spawning season with stage 3 to stage 5 gonads. The predicted percentage of mature fish at each length derived using logistic regression analysis is shown. The sample size within each 50 mm size class is shown along the top of the graph.

Table 2.7. Length at maturity (L50/L95) and 95% confidence limits derived for female Seriola hippos.

Parameter L50 (FL mm) L95 (FL mm) Estimate 831 942 Upper 95% 867 992 Lower 95% 803 902

65 Fecundity

Oocyte diameter frequency distributions in the ovaries of three reproductively developed female S. hippos (i.e. stage 4) indicate oocytes at several different stages of development (Figure 2.16). A prominent modal class at 50 – 99 μm represents oocytes at the chromatin nucleolar and perinucleolar stages. Oocytes >500 μm correspond to vitellogenic oocytes (i.e. yolk granule stage), while the oocytes with intermediate diameters were generally cortical alveolar oocytes. In each case a largely continuous overall distribution of oocyte diameters is displayed, which indicates asynchronous oocyte development and, thus, indeterminate fecundity in this species. Fecundity estimates therefore pertain to the number of eggs released per spawning event (batch) and not annual fecundity.

25

20

15

10 Frequency (%)

5

0

20

15

10 Frequency (%) 5

0

20

15

10 Frequency (%) 5

0 0 100 200 300 400 500 600 700 Oocyte diameter (μm) Figure 2.16. Oocyte frequency distributions for stage 4 ovaries of three female Seriola hippos during the spawning season.

66 A relationship between batch fecundity (BF) and FL (mm) was obtained from counts of hydrated oocytes within ovaries that were determined to be pre-spawning (stage

5a) through histology. Fecundity estimates were made for six females ranging in size from 1060 to 1200 mm FL and from 16 to 24.5 kg WW. These ranged from 51122 (±

5361) to 1472000 (± 21628) showing an increase in egg number with increasing FL and

WW of fish. The BF to FL (mm) relationship was explained with a linear equation, while the BF to WW (kg) relationship was described with a power curve:

BF = 5021 x FL – 4698076 (R2 = 0.815, n = 6)

BF = 3474 x WW 1.901 ( R2 = 0.750, n = 6)

Mortality

Age frequency distributions of S. hippos sampled in this study had a modal peak at

6 years of age (Figure 2.17). Age at full recruitment was thus considered to be 7 years and catch curve analysis was undertaken on data for 7+ and older age classes (Ricker 1975).

The resulting point estimate for the instantaneous rate of total mortality, Z, was 0.21 year-1

(Figure 2.19b, Table 2.8).

5

y = -0.208x + 5.378 R2 = 0.877 4

3

2 Ln Frequency Ln

1

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Age Class (years)

Figure 2.17. Log transformed age frequency distributions of Seriola hippos used to determine Z. Regression equation and coefficient of determination values are also shown.

67 The point estimate for the instantaneous rate of natural mortality, M, derived by refitting the relationship of Pauly (1980) was 0.40 year-1 (Figure 2.18a) a value greater than the point estimate of Z determined by catch curve analysis. The point estimate of M, derived when the maximum age of S. hippos was inserted into the refitted Hoenig relationship for fish was 0.20 year-1 (Figure 2.18a). The combined point estimate of M, derived using the method of Hall et al. (2004), which combined the separate likelihood distributions for the two estimates of M, produced a point estimate of 0.22 year-1 (Figure

2.18b, Table 2.8). The resultant posterior probability distribution for this point estimate of

M, the point estimate of Z (catch curve analysis), and the requirement that M < Z, yielded a point estimate for M of 0.16 year-1 (Figure 2.19a, b, Table 2.8) (Hall et al. 2004). The resultant 95% confidence limits for this estimate of M for S. hippos was narrower than those for M derived from refitting the relationships of Pauly (1980) and Hoenig (1983)

(Table 2.8). The point estimate of the current level of fishing mortality, determined by

Monte Carlo analysis, was 0.04 year-1 (Table 2.8).

Table 2.8. Estimates and their 95% confidence limits for total, Z, natural, M, and fishing mortality, F, for Seriola hippos in Western Australia, calculated using catch curve analysis, life history models (Pauly 1980, Hoenig 1983) and Bayesian analysis.

Z, M or F Estimate Lower Upper Method of analysis (year-1) 95% 95% Catch curve analysis Z 0.21 0.18 0.23 Refitted Pauly (1980) M 0.40 0.14 1.24 Refitted Hoenig (1983) fish equation M 0.20 0.05 0.42 Combined Pauly (1980) and Hoenig (1983) M 0.22 0.10 0.44 Combined M (Bayesian method) M 0.16 0.09 0.22 Monte Carlo F 0.04 0.00 0.12

68

0.07 (a) Hoenig (1983) 0.06 Pauly (1980)

0.05

0.04

0.03 Likelihood

0.02

0.01

0.00

0.06 (b)

0.05

0.04

0.03 Likelihood 0.02

0.01

0.00

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Natural Mortality

Figure 2.18. (a) Estimated likelihood distributions for natural mortality derived from the Pauly (1980) equation and Hoenig (1983) equation for fish. (b) Combined posterior probability distributions for M for Seriola hippos, derived from the separate likelihood distributions shown in (a).

69 for method ofHall Figure 2.19.(a) equation andHoenig’s(1983)met combined likelihooddistributionfor less thanthe catchcurve estimate of Likelihood Likelihood Z 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.00 0.01 0.02 0.03 0.04 0.05 0.06 andM 0 . (a) 00 (b) for (2004) al. for et Likelihooddistributionsfornaturalmortality for Seriola hippos . 20 Mortality (year M , from thecatchcurveanalysis for Combined analysis curve Catch (Z Propability

andtheBayesian M , assuming M from the fromthe 70 is Diet

Examination of the stomachs of S. hippos captured from the west coast of W.A. revealed that this species was not feeding while aggregated for spawning. Only 4 of the

195 specimens collected from the spawning aggregations sites had stomachs that contained food. This observation is supported by anecdotal evidence from fisherman who can only catch these fish on bait after they have been excited with lures. Food was found in 87 S. hippos stomachs outside of the spawning aggregations from fish ranging from 252 to 1370 mm FL. On the basis of the stomachs that contained food the mean stomach fullness was 5.45 ±0.26 S.E. The diet of S. hippos comprised mainly a variety of teleost fishes, which occurred in ~84% of stomachs accounting for ~74% of the overall volume, and cephalopods, which occurred in ~30% of stomachs accounting for ~26% volume

(Table 2.9). Although S. hippos consumed a wide assortment of fishes, Pilchard

Sardinops sagax was the most common contribution to the overall diet by both volume

(18%) and occurrence (24%) (Figure 2.20). Other important teleost taxa in the diet of S. hippos were a group of demersal species from the genus Centroberyx (8.3%V, 6.9%F) and the semi pelagic Yellowtail Scad Trachurus novaezelandiae (7.1%V, 6.9%F).

Squid (Order Teuthida) were the most common cephalopods consumed by S. hippos contributing 17% by volume to the overall diet of this species (Table 2.9, Figure

2.20). This prey category comprised both Southern Calamari Sepioteuthis australis

(Loliginidae) and Gould’s Nototodarus gouldi (Ommastrephidae), which were combined into the single category because distinguishing characteristics were often not discernable due to digestion. Cuttlefish of the genus Sepia were also an important prey to the overall diet of S. hippos (Figure 2.20) found in over 10% of non-empty stomachs. The results of this study only offer a general description of the diet of S. hippos as sample size precluded seasonal or geographical comparisons, however, these investigations clearly show that S. hippos are carnivorous feeders consuming both pelagic and demersal prey.

71 Table 5.7. Percentage volumetric contribution (%V) and percentage occurrence (%F) of the different prey items found in Seriola hippos (252 mm to 1370 mm FL) captured in the West Coast Bioregion. The number of stomachs that contained food = 87.

Percentage Percentage Prey Volumetric Occurrence (%F) Contribution (%V)

Cephalopods 25.7 29.9 Squid (Teuthida) 17.4 19.5 Cuttlefish (Sepia sp.) 6.6 10.3 Octopus (Octopus sp.) 1.7 1.2

Crustaceans 0.1 1.2 (Penaeidae) 0.1 1.2

Teleosts 74.2 83.9 Pilchard (Sardinops sagax) 17.8 24.1 Red/Swallowtail Snapper 8.3 6.9 (Centroberyx sp.) Yellowtail scad (Trachurus 7.1 6.9 novaezelandiae) Scaly Mackerel (Sardinella 4.5 3.5 lemuru) Garfish (Hemiramphidae) 4.4 3.5 Goatfish (Mullidae) 3.9 5.8 Leatherjacket 3.7 5.8 (Monacanthidae) Grinner (Synodontidae) 3.6 3.5 Skippy (Pseudocaranx 1.9 1.2 dentex) Gurnard (Triglidae) 1.3 1.2 Wrasse (Labridae) 0.8 2.3 Whiting (Sillago sp.) 0.2 1.2 Unidentified 16.7 36.8

72 25 1-Yellowtail scad (T. novaezelandiae) 2-Goatfish (Mullidae) 3-Leatherjacket (Monacanthidae) 4-Scaly Mackerel (S. lemuru) 20 5-Garfish (Hemiramphidae) Pilchard (S. sagax) 6-Grinner (Synodontidae) Squid (Teuthida) 7-Wrasse (Labridae) 8-Skippy (P. dentex) 9-Octopus (Octopus sp.) 15 10-Gurnard (Triglidae) 11-Whiting (Sillago sp.) 12-Prawn (Penaeidae)

Red Snapper 10 (Centroberyx sp.) Abundance (%V) 1 Cuttlefish (Sepia sp.) 4 5 5 2 6 3 8 9 10 7 0 11 12 0 5 10 15 20 25 30 Occurrence (%F)

Figure 2.20. Prey items of Seriola hippos from the west coast of Western Australia.

2.4 Discussion

Age and Growth

Difficulty in age determination has been noted in other studies of Seriola species, and this is reflected in the variety of structures used for aging, such as scales, vertebrae, opercular bones, dorsal spines and otoliths (Baxter 1960, Gillanders et al., 1999a,

Thompson et al. 1999, Kozul et al. 2000, Manooch and Potts 1997a, b). However, although delineation of opaque zones was sometimes difficult to detect in S. hippos, monthly marginal increment analysis demonstrated seasonal development of the otolith margin and the formation of a single opaque zone annually in the otoliths of this species.

The recapture of a S. hippos individual subsequent to an injection of calcein also validated the annual formation of annuli in the otoliths of this species. The similarity of predicted recapture lengths and observed recapture lengths in addition to comparisons of actual growth and modelled growth further verified the age estimation techniques used in this study. Thus, as with yellowtail kingfish S. lalandi from the waters of New South Wales

(Stewart et al. 2004) and S. dumerili from the Atlantic (Manooch and Potts 1997a, b,

73 Thompson et al. 1999), it is appropriate to use the number of annuli in sectioned otoliths to age S. hippos.

The formation of the opaque zone in the otoliths of S. hippos occurs during late spring/summer (November – January) with development of a new translucent zone occurring during autumn. This is later than for other species in south-western Australia such as West Australian Dhufish Glaucosoma hebraicum (Hesp et al. 2002), Mulloway

Argyrosomas japonicus (Farmer et al. 2005), Silver Trevally Pseudocaranx dentex

(Farmer et al. 2005) and Australian georgianus (Fairclough et al. 2000), in which the delineation of the annual opaque zone occurs in spring. The processes that govern the deposition of the opaque zones and annuli in otoliths have been the source of considerable discussion (Beckman and Wilson, 1995). Annulus formation has been related to various factors resulting from interactions between internal and/or external cycles such as somatic growth, annual migrations, temperature and spawning (Beckman and Wilson 1995; Tserpes and Tsimenides 1995; Franks et al. 1999; Cappo et al. 2000;

Ewing et al. 2003). Trends exhibited in the timing of annulus formation in S. hippos correspond with the summer spawning period, which is generally associated with large- scale migration (see Chapter 3), rapid development of the gonads and a substantial decrease in feeding activity.

The formation of annuli in the otoliths of S. hippos is paralleled by other reef- associated pelagic species, including Cobia, Rachycentron canadum of the monotypic family Rachycentridae, a sister taxon to Carangidae (Reed et al. 2002). Franks et al.

(1999) reported that annulus deposition in the otoliths of R. canadum from the Gulf of

Mexico occurred during summer and was more related to the northern migration of that species during spring than to its summer spawning activities. It is therefore likely the annulus deposition in S. hippos is related to both spring/summer migration and spawning.

It is also possible that somatic growth slows considerably during this time as fish invest

74 considerable energy towards migration and gonadal development. Formation of the translucent zone, therefore, does not occur until summer spawning activities cease and feeding and somatic growth rate increases. This notion is also supported by the fact that calcein injected into a S. hippos individual in February was incorporated into the early stage of a translucent zone within the otoliths.

The growth of juvenile S. hippos is rapid with this species reaching minimum legal length for retention (MML) of 600mm TL within the second year of life. Fast growth continues during the first 5 years of life, after which it slows down as individuals of this species become sexually mature. Once maturity is reached female S. hippos grow at a faster rate and attain a larger size than males. This pattern of sexual dimorphism is common in many species of fish (see Parker 1992 and references therein) and has also been recorded in other species of carangids, such as Blue Spotted Trevally Caranx bucculentus (Brewer et al. 1994), and Caranx hippos (Kishore and Solomon

2005), and also R. canadum (Brown-Peterson et al. 2001). However, in a study on the closely related Seriola dumerili from the Gulf of Mexico, Thompson et al. (1999) found no differences in the growth rates of females and males. Likewise, Stewart et al. (2004) found no differences in rates of growth of male and female S. lalandi from the waters of

New South Wales. Difference in growth rates and maximum size of female and male S. hippos may reflect different selection pressures acting on the body size of the two sexes.

A larger size in female S. hippos is favoured because it increases fecundity, whereas male size is most likely driven by male-male competition for females and egg fertilization

(Parker 1992). Intense competition between males for a larger female during group broadcast and paired spawning events has been reported in other carangid species, including S. dumerili and S. lalandi which do not display different in growth rates

(Graham and Castellanos 2005; Moran et al. 2007). However, although Thompson et al.

(1999) reported no differences in rates of growth between female and male S. dumerili,

75 these authors did observe sex related differences, in that females grow larger than males, and attributed these findings to age related differential mortality because males die younger. This is not the case for S. hippos where the relative proportion of males to females remains about the same for each age class.

The maximum age recorded for S. hippos in this study was 29 years. This is considerably higher than the maximum ages published for any other Seriola species. The oldest S. dumerili recorded was 17 years from the North Atlantic Ocean (Manooch and

Potts 1997b), whilst Thompson et al. (1999) recorded a maximum age of 15 years for that species in the Gulf of Mexico. In eastern Australian waters S. lalandi has been documented to attain an age of 21 years (Stewart et al. 2004). The higher longevity of S. hippos compared to other Seriola species may be attributed to this species living in cooler higher latitude waters (Heibo et al. 2002). Similarly, the oldest carangid thus far studied is Pseudocaranx dentex from temperate New Zealand waters that was reported to reach an age of 46 years (James 1984).

Juvenile Ecology

Juvenile S. hippos < 100mm in length were captured in surface waters underneath jellyfish or associated with drifting detached seagrass and brown algae, such as Posidonia,

Sargassum and Ecklonia species, in waters ranging in depth from 5 to 150m. Although individuals of this size were often encountered under a single piece of flotsam or clump of seaweed, they were more commonly encountered associated with large drifting mats of debris or windrows. Such floating structures increase the complexity of the pelagic environment and provide refuge from predators and supply a direct food source to many species of both pelagic and inshore, juvenile fish (Helfman et al. 1997), and are considered to be an integral component of many pelagic food webs (Rooker et al. 2006).

Pelagic sargassum mats have been described as important habitat for other small juvenile

76 Seriola species, such as S. dumerili in the Gulf of Mexico (Wells and Rooker 2004),

Japanese Amberjack S. quinqueradiata in the East China Sea (Sakakura and Tsukamoto

1997) and Pacific Ocean (Uehara et al. 2006), and Almaco Jack S. rivoliana and Banded

Rudderfish S. zonata in the North Atlantic Ocean (Moser et al. 1998). Although juvenile sampling was opportunistic, it is likely that small juvenile S. hippos were more commonly encountered in offshore areas because of the proximity to the deep water spawning aggregations sites. Similarly, Wells and Rooker (2004), in a study on the distribution of juvenile S. dumerili associated with drifting sargassum, found a greater abundance of juveniles (<100mm) in offshore zones than in inshore areas and attributed the observed spatial patterns to the offshore spawning habits of that species. Although the age of small juveniles was not validated in this study, as was done for S. dumerili by Wells and Rooker

(2004), it is possible that small juvenile S. hippos (up to 65 mm FL) remain in the surface layers for up to 59 days. Larger juvenile S. hippos (200 – 500 mm FL) were found to inhabit inshore areas that contain structures such as rocks, reef, pylons or jetties and where the water depth is <20 m. Whilst individuals throughout the size range were encountered in inshore areas, the sampling regime and tagging suggest that this species moves into deeper water as it increases in size. For instance, a 250mm juvenile S. hippos tagged at the mouth of the Swan River was recaptured as a sub-adult fish (790 mm TL) 17 km off the mainland coast (Australian National Sportfishing Association unpublished data).

Reproduction and Spawning

The mean monthly GSI’s of female S. hippos, together with the changes in ovarian stages throughout the year and the prevalence of fish with spawning ovaries (stage 5) in

November to March, demonstrate that this species spawns during late spring and throughout summer. Similarly other large Seriola species have been reported to spawn during late spring and summer, such as S. dumerili (Harris 2004) and S. lalandi from

77 eastern Australia in which peak gonad activity was observed in December (Gillanders et al.1999b).

The presence of post-ovulatory follicles in female gonads that were not fully spent, and the fact that oocytes of all stages were present in the ovaries of spawning females, confirms that S. hippos is a serial spawner with indeterminate fecundity. Thus, the potential annual fecundity of S. hippos is not fixed prior to the commencement of spawning and females of this species release more than one batch of eggs during the spawning season (Hunter et al. 1985). Fecundity estimates provided by this study show that S. hippos is highly fecund, particularly given the fact that the highest batch fecundity estimate of almost 1.5 million eggs was obtained from a female that was well below the maximum size attained by this species. The mean relative fecundity calculated for S. hippos of 57.4 eggs/g ovary-free body weight is comparable to other pelagic species for which such data are available. This includes, R. canadum, for which Franks et al. (1999) estimated a mean relative fecundity of 53.3 egg/g ovary-free body weight. Similarly,

Farley and Davis (1998), determined the relative fecundity of Southern Bluefin Tuna

Thunnus maccoyii to be 57 eggs/g, while Schaefer (1996) a found a relative fecundity of

68 eggs/g in Yellowfin Tuna Thunnus ablbacares.

As mentioned earlier, the annual aggregations of S. hippos that occur west of

Rottnest Island are for spawning purposes. Domeier and Colin (1997) define a spawning aggregations as “a group of conspecific fish gathered for the purpose of spawning, with fish densities and numbers significantly higher than those found in the area of the aggregation during non-reproductive periods”. More specifically Domeier and Colin

(1997) suggest that an increase in density of spawning fish greater than 3-fold constitutes a spawning aggregation. Although no attempt was made during the present study to determine actual numbers of S. hippos at the spawning sites the use of simple underwater video equipment together with commercial grade sounding devices revealed thousands of

78 S. hippos individuals in sites where very few are encountered during the non-spawning period (A. Bevan pers. comm., A. Rowland pers. obs.). For instance, observations made by sounding equipment revealed that the S. hippos aggregation on the Outer Patch commonly covered a circular area of 1250 m2 (i.e. diameter of 40m) with a high density of fish rising 40 m off the bottom. This, together with video revealing a tightly pack school of at least one individual per cubic meter, allows for a rudimentary estimate of 16 800 individuals at the aggregation site at any one time throughout the spawning period (i.e.

⅔πr3, where r is 20 = 16 800 m3). At many times during the spawning season the densely packed Outer Patch aggregation actually covered a much greater area of up to 1875 m2

(i.e. 75 m x 25 m, perhaps due to fish orientating over the length of the wreck) with fish rising 20 m off the bottom. These figures give an upper estimate of 37 500 individuals at the aggregation site during these times (i.e. 25 x 70 x 20 = 37 500 m3). Interestingly, these estimates are in close agreement to those determined during bioacoustic surveys over the

Outer Patch using both single-beam and multibeam echosounders (M. Parsons, Curtin

University, pers comm.). Thus, since the timing of ovarian development coincides with the influx of large numbers (i.e. > 3-fold increase) of this species into the area from

October to March, these groups can indeed be referred to as spawning aggregations.

Although many carangid species aggregate for spawning (Honebrink 2001, Graham and

Castellanos 2005), accounts of spawning aggregations in Seriola species are rare.

However, following the examination of five individuals which revealed hydrated eggs,

Sala et al. (2003) confirmed a spawning aggregation of S. lalandi which comprised approximately 80 fish near an island in the central Gulf of California. Anecdotal information suggests that the larger members of the genus often spawn offshore in large groups. For example, Baxter (1960) described observations by a purse seine fisherman that witnessed S. lalandi spawning 70 miles off the coast of California in 52 fathoms (110 m), at which time hundreds of individuals swam in short circles near the surface releasing

79 masses of eggs and sperm. Gillanders et al. (1999b) also suggest that S. lalandi in eastern

Australia spawn in waters well offshore.

Although this study focussed on the spawning aggregations in deep water west of

Rottnest Island, several female fish supplied by commercial fishers from the south coast were actively spawning indicating that spawning in this species is not restricted to any particular region. Furthermore, many recreational fishers report large aggregations of spawning S. hippos in shallower waters, particularly in areas south of Cape Naturaliste and west of Mandurah. During the current study S. hippos with gonads at spawning stages were caught at various depths (5-200 m) and locations (1 to 60 km from shore) in western and southern coastal waters of W.A. Tagging (see Chapter 3) has revealed that this species move between the spawning aggregation sites near Rottnest Island and that fish tagging whilst actively spawning near Cape Naturaliste move northwards to these aggregations during the spawning season. It therefore appears likely that individuals of this species migrate considerable distances along the coast of south western Australia during the spawning season and engage in spawning activities at various locations (see

Chapter 3).

Diet

Seriola hippos can best be described as an opportunistic carnivore which feeds on a variety of pelagic and demersal prey. Dietary investigation of S. hippos captured from the Rottnest Island spawning aggregation sites revealed that this species does not feed during that time. Moreover, during peak spawning most mature individuals of each sex exhibit a highly compact stomach enveloped by gonad tissue. In many cases the stomach sits within a defined depression within the gonad tissue to the point where expansion during food intake would appear to be inhibited as the gonad occupies the vast majority of the body cavity volume. Prey consumed by S. hippos is similar to that described for other

80 large members of the genus. Baxter (1960) described S. lalandi in the eastern Pacific

Ocean as an “opportunist feeder” which consume predominantly small schooling fish species such as Sardinopes caerulea, and the Northern Engraulis mordax, and also had an important invertebrate content made up of squid (Loligo sp.) and pelagic red swimming (Pleuroncodes planipes). Similarly, Andaloro and Pipitone

(1997) revealed S. dumerili in the central Mediterranean Sea to be an essentially piscivorous predator that consumed both pelagic and demersal teleosts as well as cephalopods. Indeed the dietary composition of S. hippos closely reflected that of S dumerili found by Andaloro and Pipitone (1997) where fish (primarily Spanish

Sardinella aurita and European Pilchard Sardina pilchardus) occurred in ~80% of stomachs (cf ~84% for S. hippos) and cephalopods (including squid of the genus Loligo and European Common Cuttlefish Sepia officinalis) occurred in ~27% of stomachs (cf

~30% for S. hippos). Thus, S. hippos occupies a high trophic level and can be considered as an important top end predator in the coastal habitats throughout its distribution.

Mortality

As the point estimate for natural mortality derived for S. hippos from Pauly’s

(1980) method was far higher and thus inconsistent with the point estimate derived for total mortality, Z, from catch curve analysis, it is likely that the method of Pauly (1980) provided an erroneous value for M. The fact that the Hoenig (1983) point estimate is slightly less than the value of Z would indicate that current fishing mortality is light.

However, it should be recognised that the confidence intervals for this (and Pauly’s method) were very wide.

The Bayesian approach of Hall et al. (2004) is thought to provide a more reliable estimate of M, as it combined the likelihood distributions associated with each of the point estimates of M and the point estimate for Z, i.e. was based on greater information through

81 employing several probability distributions for Z and M. This approach always ensures M is less than Z (Hall et al. 2004, Mant et al. 2006). As the Bayesian approach incorporates the distribution values for M derived by the Pauly (1980) equation, an erroneously high estimate for this variable, bias will be incorporated into this estimate of M. However, when the estimate of M from the Pauly (1980) method was combined with that of

Hoenig’s (1983) method, the resultant value was very close to the original Hoenig (1983) estimate, i.e. the Pauly estimate has little influence on the overall estimate. Nevertheless, the estimate of fishing mortality, F, should be treated with some caution. The low point estimate for F of 0.04 year-1 suggests that S. hippos is not currently subjected to heavy fishing mortality. This agrees with the general perception by recreational anglers that S. hippos is primarily a catch-and-release sportfish rather than a table fish. Although about

80 – 100 tonnes of S. hippos are captured by commercial fishers annually, given the large individual size of this widely distributed and apparently abundant species (see Chapters 1 and 3), it is possible that these catches represent a minor proportion of the overall population.

Management

The fork length at which female S. hippos typically attain maturity (L50) (831 mm) is equivalent to a total length of 888 mm, thus exceeding the minimum legal length of retention (MML) of 600 mm TL. Therefore, many S. hippos may get harvested prior to reproduction. Gillanders et al. (1999b) reported a very similar length at 50% maturity for females of 834 mm FL in the closely related S. lalandi from the waters off New South

Wales. Although the minimum legal length of S. hippos is over 250 mm less than the length at maturity, recent reductions in the maximum daily recreational bag limit of this species from four to two per fisher, along with the fact that S. hippos is generally only targeted by sportfishers for the purpose of catch-and-release suggest that there is no

82 urgency to reduce size limit to match length at maturity. Evidence that S. hippos has a high reproductive output and an offshore spawning and migratory habit that is likely to exclude many fish from the fishery, may also account for the fact that current fishing pressure does not appear to have a marked impact on the numbers of S. hippos. Certainly, the current study and the advice of many fishers indicate that stocks of this species are healthy, however, mortality associated fishing activities should be closely monitored into the future to ensure that this species is not overexploited.

83 Chapter 3

Movement and migration of Samson fish, Seriola hippos Günther 1817, in Western Australia

3.1 Introduction

Dingle (1996) classified movement behaviours into three main types, namely, station keeping, ranging and migration. Station keeping movements comprise the simplest forms that are directly related to activities that exploit available resources to promote growth and reproduction. These movements take place within the home range of an animal and may include searches for food and shelter or actions defending territory.

Ranging behaviours are similar to those of station keeping as they are resource directed movements, but are characterised by departure from the current home range and usually include a specifically exploratory component, such as movements over an area to find new resources or territory (Dingle 1996). Ranging behaviour generally ceases when a suitable habitat patch is located by the animal, for instance, a young animal moving to find space away from its parents to avoid competition (Dingle and Drake 2007). In contrast, migratory behaviour is characterised by the undistracted movement of an animal from one

84 habitat to another and is normally associated with suppressed or suspended responses to resources or home ranges whilst in transit.

Migration is a biological phenomenon that transcends taxa, form and environment

(Dingle and Drake 2007). The definition of the term migration has received considerable attention from many biologists studying a variety of organisms, including , fishes, birds and mammals (see for example Dingle 1996, Dingle and Holyoak 2001). The common viewpoint reached by many authors, is that migration generally encompasses two main concepts, namely, the behaviour of individuals (i.e. its mechanisms) and the ecological consequences applying to populations (i.e. its function). In a review of various aspects of migration across taxa Dingle and Drake (2007) state that, like other biological phenomena, migration is best defined in terms of natural selection as it is selection pressures that have formed the behaviours that distinguish migration from other kinds of movement. Thus, the definition must be specified in terms of individuals. The most widely accepted definition that applies across taxa and allows an objective distinction between migration and other forms of movement appears to be that of Kennedy (1985):

“Migratory behaviour is persistent and straightened-out movement effected by the animal's own locomotory exertions or by its active embarkation on a vehicle. It depends upon some temporary inhibition of station-keeping responses but promotes their eventual disinhibition and recurrence.”

Migration for the purpose of spawning or breeding is common and known to occur in a variety of taxa, including (e.g. Ruello 1975), fish (e.g. Leggett 1977), amphibians (e.g. Pilliod et al. 2002), reptiles (e.g. Parmenter 1983), birds (e.g. Berthold

2001) and mammals (Matthews 1978, Stevick et al. 2004). In Chapter 2 it was demonstrated that the large deep water S. hippos aggregations that occur annually west of

85 Rottnest Island during late spring, and which persist until early autumn, are spawning aggregations. In addition, such a large annual, temporary influx of S. hippos individuals suggests that members of this species undertake spawning migrations. Annual spawning migrations in sea fishes (termed oceanodromous) are relatively common and are considered important life history events that function to connect individuals from large areas to specific spawning grounds (Arnold 2001).

Aggregation spawning is one of the most striking reproductive strategies that have evolved among fishes to maximise spawning success. This strategy, which often occurs at predictable times and locations, may enhance the capacity of individuals to select mates, synchronise spawning and optimise survival of offspring (Thresher 1984, Colin and

Clavijo 1988). In many species tens to thousands of individuals may travel long distances from residential habitats to assemble at specific sites to spawn in pairs, harems or large groups (Domeier and Colin 1997). Some of the best known examples include species of the families Serranidae (Rhodes and Sadovy 2002, Nemeth et al. 2007), Lutjanidae (Carter and Perrine 1994, Burton et al. 2005, Heyman et al. 2005) and Carangidae (Graham and

Castellanos 2005, Chapter 2). Aggregations consisting of fish that have travelled relatively long distance and persist at the aggregation site for days or weeks (i.e. the reproductive season) are termed transient spawning aggregations, whilst those involving fish that have travelled short distances and persist for minutes to hours, which often occur more frequently, are termed resident spawning aggregations (Domeier and Colin 1997).

Studies of fish aggregating and migratory behaviour typically involve tracking of individual fish. Cooperative tagging programs, whereby fisheries agencies provide tags to volunteer recreational anglers, are cost effective methods to study fish and fish populations (Ortiz et al. 2003). In addition to obtaining information on stock size, growth, mortality and stocking success, one of the most common scientific reasons why such programs are initiated is to collect information on species migrations and stock structure

86 (e.g. Moran et al. 2003). Cooperative tagging programs can also be undertaken to increase public awareness of research. Fish tagging is seen to have an intrinsic political value as it is a highly visible feature of research which the public, in general, recognise and identify with (Kearney 1989). Such tagging programs can also increase the conservation ethic among anglers as direct research involvement positively influences fishing practices and can lead to increased community stewardship of the resource (see Chapter 4). Fish tagging by recreational anglers is often undertaken exclusively for social reason such as to justify the capture and release of, as opposed to the killing of, species with high emotive value, such as marlin (Kearney 1989).

In light of both the scientific and social considerations of fish tagging, a large scale collaborative research project titled ‘Samson Science’ was initiated during the early phases of this study. The intention of this project was to engage the Western Australian recreational fishing industry and community in the research of this popular sportfish (see

Chapter 1). The main objective was to tag and release a large number of S. hippos from the Rottnest Island aggregations, using a number of strictly controlled capture, handling and release methods, with accurate and detailed recording of data. These data, along with subsequent recapture information, were then to be used to:

1) describe movement/migration patterns of S. hippos to and from the aggregations,

2) describe the S. hippos sportfishery and fishing behaviour (Chapter 4),

3) develop the handling protocols for the sportfishery (Chapter 4), and

4) estimate the numbers of S. hippos at the aggregations sites.

The last objective was discounted in the early stages of this study since short-term recaptures revealed that any estimates of fish numbers would be greatly biased by fish movement patterns (Schwarz and Seber 1999, and see discussions below).

87 This chapter aims to test the hypothesis that the S. hippos spawning aggregation can best be described as transient spawning aggregations sensu Domeier and Colin (1997), i.e. individuals may travel large distances to specific sites annually to join these aggregations. In addition, this chapter aims to establish the length of stay at the spawning sites of individuals and to determine whether S. hippos move between different spawning sites during the same spawning season. The results presented here, together with biological understanding developed in Chapter 2, will also be used to consider how natural selection may acted to produce observed movement patterns and to develop hypotheses pertaining to the life history strategies undertaken by S. hippos.

3.2 Materials and Methods

Samson Science

A large scale collaborative research project, titled ‘Samson Science’, was initiated in 2004 to study the Rottnest Island S. hippos aggregations. This project, which involving

Murdoch University and the Department of Fisheries Western Australian (DoFWA) incorporated a large community extension component to directly engage the Western

Australian recreational fishing sector. The initial research aim of Samson Science was to tag 3000 S. hippos at the aggregations near Rottnest Island in a 3 week period (January

2005) to estimate fish numbers, determine movement patterns, assess effects of fishing and to develop protocols for handling fish. However, this initial tagging period was extended by two weeks to make up for days lost to poor weather.

In August 2004 the Samson Science project initiated community engagement with extensive advertising in media aimed at recreational fishers such as local fishing magazines, radio programs and fishing columns in newspapers. A database was established, for ongoing communication and project planning, containing the names and contact details of volunteer fishers together with information on their capacity for research

88 involvement. A regular newsletter was distributed (electronically and through the post) to research volunteers to keep them informed of project developments and future training and information sessions. A popular Western Australian recreational fishing website, Western

Angler, established a forum specifically for the Samson Science project to aid in communication of project developments and preliminary results to the local recreational fishing community. This forum also facilitated coordination of, and communication between, volunteers and participants involved in the project.

A series of seminars and workshops was held in Perth and other locations to raise awareness about the study and to recruit and train volunteer fishers, taggers and support staff. In each tagging workshop fishers were shown how and where to apply tags using fish previously collected for the biological component of this study (see section below for tagging methods). During the workshops, the importance of the research, the ethical considerations of fishing and the need for rigour in the collection of data were explained and discussed. At all stages the importance of minimising trauma to the fish was stressed, whilst also noting that anglers should continue their normal fishing practices.

Support was received from local businesses that provided rewards for participants such as fishing tackle, restaurant vouchers, boat training courses and fuel vouchers. These rewards, together with Samson Science T-shirts, were provided as incentives to project participants, particularly as each fishing trip was a relatively costly exercise. Prizes, provided by industry, were given for particular feats, such as the most complete datasheets and largest the fish tagged. Feedback of results and information were provided to stakeholders via the ‘Samson Science’ newsletters and a forum on the Western Angler website, via workshops, radio interviews, and reference in other media. Data entry and coordination of tagging events was assisted by volunteer fisheries liaison officers from

Department of Fisheries Western Australia. Fremantle Sea Rescue provided radio

89 coverage, and the charter vessel ‘North Star 2’ assisted with at-sea coordination of vessels and tagging.

The Samson Science tagging project completed in January 2005 was widely acclaimed a successful collaboration between scientists and recreational fishers. Although the Samson Science project was initially intended as a single event there was considerable interest amongst the recreational fishing community to contribute further in this research.

Given the success of this collaboration and the encouragement of recreational anglers and

Recfishwest (the peak body representing recreational fishers in Western Australia), additional funding was received from the Fisheries Research and Development

Corporation for a second season extension (2005-06). Samson Science 2 had smaller metropolitan tagging goals and a new focus on tagging in southern areas. The focus on southern waters occurred because virtually all recaptures of fish tagged at the Rottnest aggregations were south of the tag location (see Results). Samson Science 2 also required an intensive period of seminars and workshops.

As noted above, ongoing extension of project developments and findings was continued through the regular newsletters and website forums as well as post event results seminars and media releases. Extension through website forums, telephone conversations and meetings with core participants was continued through the life of the project (over 3 years).

Tagging

Anglers were provided with a tag kit per boat and data sheets. In addition to latitude and longitude, angler, tagger, tag number and fish length, the data sheets also included sections for how high the fish was lifted from the water, fishing method (i.e. jigs or baited hooks), line class, hook position, lift method, revive (release) method, release

90 condition and a comments section where anglers could provide other details, e.g. was the fish bleeding, had the fish been previously tagged, etc. (Appendix).

In January 2005, S. hippos were tagged at three spawning aggregation sites in waters west of Rottnest Island in a large scale intensive tagging program. During this time fish tagging was focused on three specific sites known as North Barge, Outer Patch and

South Barge (Figure 3.1). All fish were caught by hook and line using either bait or artificial metal jigs. Once a fish was landed, the fork length was measured to the nearest cm and a nylon headed, single barbed dart tag (Hallprint PDA, 120mm long, 3mm diameter) with a distinctive identification number was inserted to lock under the pterygiophores of the anterior section of the 2nd dorsal fin. The fish was then released by the spearing method (see Chapter 4) or, if deemed necessary, the use of a release weight

(see Chapter 4) in an effort to ensure descent back to the school. Any fish that did not release successfully and floated on the surface was gathered for another release attempt.

Figure 3.1. Seriola hippos aggregation sites the metropolitan region where tagging was conducted.

91 Tagging involving researchers and recreational fishers was also conducted between

May 2005 and September 2006 in the south west region of Western Australia. Although, fishers were recruited to tag S. hippos in any coastal waters between Mandurah and

Margaret River, tagging was concentrated in areas near Dawesville and Cape Naturaliste

(Figure 3.2). A four week intensive tagging event involving researchers and recreational fishers was then again undertaken in the second summer (November 2005) within the metropolitan area. Tagging during this season was again focused on the Rottnest aggregations as well as a north metropolitan aggregation known as the Hillarys Barge

(Figure 3.1). The objective of this tagging event was to investigate the length of stay at, and movement between, the metropolitan aggregations (particularly linkage between north and south metropolitan aggregations). Furthermore, undertaking an intensive tagging event during this time increased the chance of recapturing fish tagged during the previous month in the Southern regions. In addition to the intensive tagging event of November

2005 a prominent charter boat within the fishery, North Star 2, and a group of recreational fishers who regularly fished the aggregations, continued tagging S. hippos until March

2006.

A third summer (Nov 2006 – March 2007) tagging project was conducted with

North Star 2 and the aid of a select group of anglers who had, during the previous summers, built up a deep interest in the project and had proven themselves as reliable taggers.

Seriola hippos tagged previous to the current project at the spawning aggregations west of Rottnest Island by the charter boat North Star 2 as part of the Australian National

Sportfishing Association’s Westag program (n = 2265) are also included in the present study.

92

Figure 3.2. Map of the south-west region of Western Australia where tagging was conducted. Areas marked blue represent areas of most tagging effort. Red areas contain the spawning aggregations near Rottnest Island.

During the tagging of S. hippos at the spawning aggregation sites the sex of some individual fish was determined by research staff prior to release by inserting a cannula 20 to 30 mm into the urogenital pore. The cannula used was a flexible vinyl tube 30-40 cm long with a small bore (2mm inside diameter). Insertion of the tube enabled eggs or sperm to be extracted, in some cases a slight vacuum was applied by sucking on the end of the tube. Alternatively, when spawning was at its peak, slight pressure applied to the abdomen would generally expel eggs or sperm.

Time at liberty was calculated as the difference in days between the date of release and the date of recapture. Distances travelled by individual fish between release and recapture sites were measured by the most direct water route. The relationship between

93 distance moved and time at liberty for recaptured S. hippos was investigated by use of the

Spearman rank correlation coefficient (rs).

When attempting to use tag and recapture data to make inferences about patterns of movement and growth rates the following assumptions are generally necessary (Kearney

1989, Gillanders et al. 2001):

1) The behaviour of tagged fish is same as that of the non-tagged population,

2) Tagged fish are released into, and fully mixed with, the mainstream population, thus, the tagged sample is representative of the population as a whole,

3) Patterns of recaptures represent distribution patterns of tagged fish and therefore the population as a whole,

4) Tagging does not effect growth,

5) The time and location of tagging and tag recovery is properly recorded and,

6) The lengths of individuals is recorded without bias at the time of tagging and tag recovery .

The extent to which any of the above assumptions are violated will be discussed in light of the results of this chapter.

Leeuwin Current

Movements and migrations of many species of fish have been linked to prevailing ocean currents (Hutchings et al. 2002). In Western Australia the Leeuwin Current has been shown to greatly influence the movement, distribution patterns and recruitment of numerous marine organisms (Lenanton et al. 1991, Fairclough et al. 2000). Pearce and

Phillips (1988) hypothesised that coastal sea level at Fremantle could be used as an index of the strength of this southward flowing current. Since then Fremantle sea level data have been used to represent Leeuwin Current strength by a number of fisheries researchers

(see for example Caputi et al. 1996). This approach was later justified by Feng et al.

(2003) by confirmation of a linear relationship between coastal sea level deviation at

94 Fremantle and the Leeuwin Current transport. Data for mean monthly sea level at

Fremantle between November 2001 and March 2006 were obtained from the National

Tidal Centre, Australian Bureau of Meteorology.

3.3 Results

A total of 9769 S. hippos were tagged at spawning aggregation sites near Rottnest

Island (Figure 3.1) between November 2001 and January 2007, of which 203 (2.1%) were recaptured (Table 3.1). All fish were tagged in the months from September to April in waters ranging in depth from 95 to 120 m. Fish tagged at the spawning aggregations ranged in length from 550 to 1600 mm FL and were recaptured after being at liberty for between 0 to 1795 days (almost 5 years), with the average time at liberty of 385 days (±

25 days S.E.). Of the total number of S. hippos tagged at the spawning aggregations, 7504 were tagged and released during the current study of which approximately 90 % were tagged at the Outer Patch and North Barge (Table 3.2). For fish in which sex was determined prior to release (n = 1762), recapture rates of females to males was 2:1 (n =

30).

Table 3.1. Release and recapture information for Seriola hippos tagged (1) at the spawning aggregations near Rottnest Island between November 2001 and January 2007 and (2) for fish tagged in the south-west region of WA between May 2005 and September 2006. Number recaptured and percentage recaptured refer to overall recaptures of fish tagged during corresponding periods.

Number Number Percentage Years / Seasons / Region Tagged Recaptured Recaptured 2001-2004 (Prior to Samson Science) 2265 77 3.4 Samson Science 1 (2004/05) 3163 73 2.3 Samson Science 2 (2005/06) 3322 49 1.5 2006-2007 (Post Samson Science) 1019 4 0.4 Spawning aggregation total 9769 203 2.1 Samson Science South-west (2005/06) 518 26 5.0 Overall 10287 229 2.2

95 Table 3.2. Release and recapture information for Seriola hippos tagged at each of the spawning aggregations near Rottnest Island between November 2004 and January 2007. Number recaptured and percentage recaptured refer to recaptures of fish tagged at each location.

Number Number Percentage Release Location Tagged Recaptured Recaptured Outer Patch 4166 66 1.6 North Barge 2612 51 2.0 South Barge 386 8 2.1 Hillarys Barge 340 1 0.3 Total 7504 126 1.7

In addition to tagging at the spawning aggregations, a further 518 S. hippos were tagged with the aid of recreational fishers in coastal waters of south-west Western

Australia between Dawesville (32° 36.00’S, 115° 37.70’E) and Margaret River (33°

58.00’S, 114° 59.00’E) between May 2005 and September 2006, of which 26 (5.0%) have been recaptured (Table 3.1). The tagging effort in the south-west region was widely distributed with the majority of S. hippos being tagged near Dawesville in waters ranging in depth from 13 to 28 m, and near Cape Naturaliste (33° 32.00’S, 115° 01.00’ E) in waters ranging from 38 to 55 m depth. Fish tagged in this region ranged in length from

380 to 1480 mm FL. Recaptured fish tagged in this region had been at liberty for between

4 to 743 days. One volunteer recreational fisher within this south-west region tagged 187

S. hippos with-in a relatively small area (< 20 km2) in the inshore waters west of

Dawesville. The recapture rate of these 187 fish was 8.6% and included three fish that were recaptured twice within the same area.

Most of the recaptured S. hippos originally tagged at the metropolitan spawning aggregations were recaptured either at the metropolitan spawning aggregations (56%, n =

113) or between 10 and 2500 km south and east (41%, n = 83) of these aggregations

(Table 3.3, Figure 3.3). Of the 113 S. hippos recaptured at the metropolitan spawning aggregations 109 (96%) were recaptured within the spawning period. Of these 30 were

96 captured during the same spawning season and 79 were recaptured during subsequent spawning seasons. Most S. hippos originally tagged at the spawning aggregations that were recaptured in waters to the south were recaptured during the non-spawning season

(59%, n = 49) (Table 3.4). Only 3% (n = 7) of the recaptured S. hippos tagged at the metropolitan spawning aggregations were encountered to the north of these sites, having travelled distances of between 10 and 330 km.

Table 3.3. Number of recaptured Seriola hippos by recapture location and release location between November 2001 and January 2007. Recapture Location Outer North South Hillarys South of North of Release Location Patch Barge Barge Barge Rottnest Rottnest Total Outer Patch 27 10 1 4 26 2 70 North Barge 22 37 2 4 55 4 124 South Barge 3 1 1 2 1 8 Hillarys Barge 1 1 South-west Region 2 4 20 26

KANGAROO IS.

Figure 3.3. Map of south-western Australia showing recapture locations (red dots) of Seriola hippos tagged and released at the summer spawning aggregations west of Rottnest Island (blue dot). The inset shows the distribution of this species in Australian waters and indicates the position of Kangaroo Island (two recaptures), highlighting the large area over which individuals tagged at the Rottnest aggregations were recaptured.

97 Table 3.4. Number of Seriola hippos released and recaptured at the metropolitan spawning aggregations and in the South-west region during the spawning and non- spawning periods between November 2001 and January 2007. NB. spawning period represents months of November to February.

Recapture Location Rottnest Island South of Rottnest Release Location Spawn. Non Spawn. Spawn. Non Spawn. Rottnest Island 109 4 34 49 South-west 6 0 7 13

Of the 83 S. hippos recaptured south of the spawning aggregations, 80% (or 33% of all recaptures) had moved >100 km and 41 % (or 17% of all recaptures) had moved >

500km. The mean distance travelled by fish that had moved south of the aggregations was

440 km (± 52 km S.E.). The sex of 15 of these fish was determined prior to tagging with

6 males and 9 females travelling a mean distance of 618 km (± 382 km S.E.) and 595 km

(± 241 km S.E.), respectively. No significant difference was found between the mean distance travelled by males and females (t = 0.053, P = 0.958, df = 13). Two S. hippos, one female and one male, tagged at the Rottnest aggregations were recaptured on the eastern side of the Great Australian Bight near Kangaroo Island in South Australia, a distance of over 2400 km (Figure 3.3), these fish were at liberty for 215 and 276 days, respectively.

The number of recaptures versus days at liberty for fish tagged at the spawning aggregations sites showed distinct periodicity with peaks in recaptures corresponding to the return of S. hippos to the summer spawning aggregations (Figure 3.4). The mean number of days at liberty for S. hippos recaptured at the spawning aggregation sites during the two subsequent spawning seasons was 372 (± 6.1 SE, n = 49) (median = 364 days) and

721 days (± 6.1 SE, n = 22) (median = 721), or almost exactly one or two years later. Of the fish returning in the first subsequent spawning season 63 % were recaptured at the same aggregation compared with 44 % returning in the second subsequent season. In total

98 58% of all S. hippos that returned to the Rottnest Island spawning area were recaptured at the exact release location.

20

Recaptures at spawning aggregations Recaptures away from spawning aggregations

15

10 Number of recaptures Number 5

0 0 365 730 1095 1460 1825 Days at liberty

Figure 3.4. Number of recaptures plotted against time at liberty for Seriola hippos tagged at the spawning aggregation sites west of Rottnest Island. Days at liberty are in 10 day intervals. Recaptures at, and away from, the spawning aggregations are indicated.

Thirty S. hippos were released and recaptured at the spawning aggregations sites within the same spawning season. The duration at liberty ranged from 0 to 36 days with an average of 12 days (± 1.8 days SE). This indicates that fish can stay at the aggregation sites for over a month. Although most of these fish (70 %) were recaptured at the same aggregation site, 30% (9 individuals) were recaptured at an aggregation other than the site of release, with six fish moving from the Outer Patch to the North Barge (Figure 3.5).

99

Figure 3.5. The location of Rottnest aggregations showing the number of Seriola hippos tagged and either recaptured at the same aggregation or at another aggregation (arrows indicate the direction of movement).

Movements between spawning aggregations were also considerable between spawning seasons (Table 3.3). Of the 124 recaptured S. hippos released at the North

Barge, 22 (18%) were recaptured at the Outer Patch in subsequent spawning seasons.

During the same period, 10 (14%) of the 70 recaptured fish tagged at the Outer Patch were recaptured at the North Barge and eight fish from the spawning aggregations near Rottnest

Island moved to the spawning aggregation off Hillarys (Table 3.3).

Tagging of S. hippos in the south-west region confirmed that individuals move northwards to the aggregation sites west of Rottnest Island. Of the 26 recaptured fish released in the south-west region 6 (23 %) showed northward movement > 50 km (Figure

3.6). This included four S. hippos tagged near Cape Naturaliste that were recaptured at the

Rottnest Island aggregation sites after being at liberty for between 53 and 423 days. One individual S. hippos originally released near Dawesville was recaptured after northern movement to the spawning aggregations and then recaptured for a second time after returning south to the original tag site.

100 .

Figure 3.6. Movements of Seriola hippos tagged in the South-west region of Western Australia. Green dots indicated tag location and reds dot indicated recapture location.

Many S. hippos tagged at the spawning aggregation sites showed fast southward movements. For instance, one S. hippos (1250 mm FL) released at the Rottnest Island spawning aggregations was recaptured approximately 1000 km south-east near Esperance

(34° 16.79’ S, 122° 02.38’E) after being at liberty for 26 days. This is equivalent to at least 39 km day-1. A smaller individual (1080 mm FL) was recaptured east of Bremer Bay

(34° 26.52’S, 120° 03.56’E) after 25 days at liberty and exhibited a movement rate of at least 34 km day-1. Overall, eight fish (4 %) moved at a rate > 10 km day-1 after leaving the spawning aggregations. This indicates that fish can quickly return to habitat in the south.

No correlation between distance moved and time at liberty was found for fish that were recaptured away from the spawning aggregations within a 365 day time period (R2 = 0.04, df = 40, P = 0.79) (Figure 3.7).

101 3000

2500

2000

1500

1000 Distance travelled (km) travelled Distance

500

0 0 50 100 150 200 250 300 350 400 Time at liberty (days)

Figure 3.7. Distances moved and times at liberty of Seriola hippos tagged at the spawning aggregations west of Rottnest Island. Data only shown for fish at liberty for <365 days that were recaptured away from the spawning aggregation tag sites.

Tagging suggests that S. hippos individuals exhibit similar patterns of movement and may even travel together. For instance, two individuals (1200 and 1000 mm TL) tagged at the North Barge on the same day in January 2004 were recaptured 24 and 30 days later, respectively, in the same area near Albany (35° 01.50’ S, 117° 53.50’ E).

Similarly, three individuals released near Dawesville on the same day in April 2006 were each recaptured ten kilometres north of the tag site, on the same day and by the same fisher, 80 days later. These individuals varied in length from 920 to 1210 mm TL.

No correlation existed between the time of tagging during the spawning season and the distance travelled by an individual displaying southward movement from the aggregations. For instance, individuals tagged during the month of December that were recaptured south of the tag site travelled a mean distance of 493 km (± 100 km SE) whereas those tagged in January travelled a mean of 406 (± 71 km SE). Indeed, the fish that had travelled the greatest distance (i.e. 2500 km) visited the spawning aggregations in

December.

102 Data for mean monthly sea level at Rottnest Island, used as an index of the strength of the southward flowing Leeuwin Current, shows that S. hippos first arrive at the Rottnest

Island spawning aggregations at times when the current is at its weakest (Figures 3.8). The

S. hippos spawning aggregations conclude at a time when the strength of the Leeuwin

Current is increasing.

600 110

100 500

90 400

80 300 70

200 (mm) Sea level

Number of fish tagged 60

100 50

0 40 D F A J A O D F A J A O D F A J A O D F A J A O D F A 2001 2002 2003 2004 2005 2006 Year/Month

Figure 3.8. The numbers of Seriola hippos tagged at the spawning aggregations west of Rottnest Island between November 2001 and March 2006 and the mean monthly sea level at Fremantle (an index for the strength of the Leeuwin Current) throughout that period.

3.4 Discussion

The large number of S. hippos tagged at the Rottnest Island spawning aggregations was due to a very credible volunteer effort, particularly when it is taken into account the long travel time, poor weather and exhausting task of capturing these powerful fish from over 40 kilometres out to sea in water depths of 90 and more metres. The cooperative approach to tagging undertaken during this study produced a large amount of data that could otherwise not have been gained due to cost, effort and time constraints. The pattern of recaptures of S. hippos suggests that the assumptions made in tagging studies and that

103 are related to movement were met (see assumption 1 to 3 page 95). For example, tagged fish were recaptured both on the aggregations with in the same spawning season as well as one or two years subsequently, whilst fish recaptured outside of the spawning season were often mixed with non tagged fish. Examination of Figure 2.9 (page 57) shows that tagged fish followed their expected growth trajectories, thus assumptions 4 to 6 also appear to have been met. The fact that all of these assumptions were met demonstrates that robust data, producing valuable scientific insights, can be collected by recreational fishers and charter boat operators.

Migration of Seriola hippos

This study confirmed that S. hippos undertake true spawning migrations along the southern and western coastlines of Australia. Seriola hippos movements displayed during the present study are indeed consistent with migratory behaviour given that the direction of S. hippos movement differs between the spawning and non-spawning seasons and the fact that many tagged S. hippos undertook long distance movements to and from the

Rottnest Island spawning sites (e.g. up to 2500 km). This study also revealed fast movements southward after spawning (up to at least 39 km day-1) which would infer undistracted movement, i.e. normal responses to resources are likely to be suppressed whilst in transit, and thus fill the criteria of migration behaviour defined by Kennedy

(1985). Furthermore, under the criteria of Domeier and Colin (1997) these seasonal congregations of S. hippos can be classified as transient spawning aggregations.

Although the overall recapture rate of S. hippos between 2001 and 2007 (2.2%) was similar to that recorded for S. hippos tagged in South Australian waters (2.3%)

(Hutson et al. 2007a), the rates of recapture of fish tagged at the spawning aggregations varied considerably. This is not unexpected because fish tagged in the earlier years have been at liberty for longer and therefore have had a greater chance of being recaptured.

104 The overall recapture rate of S. hippos is low compared to other marine fish for which conventional tagging studies have been undertaken within Australia, such as pink snapper

(Pagrus auratus, 7.9%, Moran et al. 2003) and yellowtail kingfish (S. lalandi, 8.2%,

Hutson et al. 2007a). The low recapture rate of S. hippos in the current study is likely to reflect the large numbers of this species at the aggregation locations and a combination of the large scale migratory movements of this species with many tagged individuals moving into southern waters where there is limited or no fishing effort. Furthermore, this species is not normally targeted by recreational fishers for eating purposes, and relatively few fishers in the southwest target this species for sport compared to the metropolitan area.

The low rate of recaptures of fish tagged at the Hillarys Barge is likely a consequence of high predation by sharks on fish as they are released. Large numbers of whaler sharks

Carcharhinus spp. frequent this aggregation site whereas these species are rarely encountered at the other three aggregations targeted. Shark predation has been documented as a major cause of post release mortality within other catch-and-release fisheries in areas where shark are abundant (Cooke and Philipp 2004). Indeed, the prevalence of sharks at the Hillarys Barge resulted in little fishing effort being continued and is reflected in the low numbers of S. hippos tagged at this location.

Distances moved by S. hippos migrating to the Rottnest aggregation sites may be relatively short, perhaps from Mandurah or closer, or they may be from as far away as

Kangaroo Island in South Australia (2500 km). Long distance movement between spawning and feeding grounds has also been reported in other Seriola species, although details are limited to anecdotal accounts (Baxter 1960, Thompson et al. 1999). In a tagging study of S. lalandi, Gillanders et al. (2001), also found large-scale movements (>

500 km) of individuals along the east coast of Australia and revealed that three fish travelled over 2000 km from New South Wales to New Zealand waters. These authors reported no evidence of migrations in S. lalandi stating that, although movements of this

105 species generally appear limited, a lack of fishing effort within particular parts of the

Pacific Ocean may bias current descriptions of the overall movement patterns of this species. Similarly, the New Zealand Cooperative Gamefish Tagging Programme, which has tagged over 10 000 S. lalandi in the country’s northern waters since 1975, has revealed mostly localised movements with 93% of recaptures having occurred within 50 nautical miles (93 km) of the release point (Holdsworth and Saul 2003). The New

Zealand tagging program, like the one in NSW, has also reported S. lalandi moving between Australia and New Zealand, albeit one individual in the opposite direction (Saul and Holdsworth 1992).

During this study, almost all recaptured S. hippos originally tagged at the aggregation sites west of Rottnest Island were caught at either the aggregation sites or in waters to the south. This indicates that the majority of S. hippos that spawn near Rottnest

Island move there from either local or southern waters. Large-scale northwards movement of this species from southern waters to the spawning aggregations is also supported by observations from fishers in the south-west region who report that, whilst S. hippos numbers remain fairly constant during winter (Jun-Aug), an increase in the numbers of large S. hippos occurs during October in the near shore waters of Geographe Bay

(Busselton). The migration patterns of fish to the north of the Rottnest Island region however remains unclear as only seven tagged fish were recaptured after northward movement. Thus, the resultant patterns of tag returns indicate that, whilst the S. hippos population of the south-west and southern coastal regions are likely to be well mixed, spatially discrete ‘southern’ and ‘northern’ adult breeding units may exist along the west coast of Australia.

Observations made at the end of the aggregating period indicate that most fish leave the Rottnest Island area quickly. None of the fish tagged in the south-west region were recaptured more than 5 km south of the original tagging location, whereas six fish

106 (26%) had moved over 50 km northwards during the spawning period. The results of this study suggest that individuals return to the same location after spawning and reside there in the general vicinity over the winter months. This idea is supported by the recapture of one individual originally released near Dawesville during the non-spawning period that moved northward to the spawning aggregations before returning again to the original tag site where it was recaptured again in the non-spawning period. Furthermore, the high recapture rate displayed in the south-west region, including the particularly high recapture rate (8.6%) by one fisher who fished within a small area near Dawesville, also indicated that this species is likely to inhabit the same reef area for most of the time between spawning events and return to that area after spawning. This recapture rate is comparable to those for S. lalandi tagged in waters off New South Wales (8%) and South Australia

(8.2%) where anglers did not target spawning aggregations and where the majority of fish were recaptured within 50 and 5 km of the release point, respectively (Gillanders et al.

2001, Hutson et al. 2007a).

Seriola hippos individuals also show strong spawning ground fidelity as many of the fish released at the spawning aggregations were recaptured at the same spawning site.

Interestingly, the temporal pattern displayed in these recaptures also revealed that most S. hippos return to the spawning aggregations at the same time in subsequent years, for example the average days at liberty of these returning fish was 372 and 721 days.

Homing to spawning grounds by individuals has long been recognised in diadromous fishes, such as numerous salmonid species (Scheer 1939) and American shad, Alosa sapidissima (Hollis 1948). However, it has only recently been confirmed in marine broadcast spawners, such as the Atlantic Cod Gadus morhua (Robichaud and Rose 2001) and Plaice Pleuronectes platessa (Hunter et al. 2003). Each of the spawning aggregations near Rottnest Island was spatially consistent over the three year period of this study due to the presence of distinct shipwrecks on an otherwise structureless substrate. Thus, return

107 migration of S. hippos individuals to the exact same spawning location (i.e. within meters) from possibly distant over-wintering habitats, which may be up to 2500 km away, suggests a very high level of navigational precision. The nature of such homing in this species is unclear. Although many such questions relating to the natural history of S. hippos are beyond the scope of this tagging study, inferences from recapture patterns and the biological understanding developed in Chapter 2 together with information on the oceanographic features of south western Australia have been used to develop hypotheses about the spawning strategies used by this species (see below).

Seriola hippos that were released and recaptured at the spawning aggregations sites within the same spawning season revealed that individuals may reside at a single spawning site for up to a month and that there is a continual exchange of S. hippos individuals at the aggregation sites as the spawning period proceeds. Thus, based on this finding (i.e. an individual stay of four weeks) together with the conservative estimated number of individual fish within the Outer Patch aggregation made in Chapter 2 (i.e. 16

800 to 37 500), a crude and very conservative estimate is that between 67 000 and 150 000 individuals visit this particular aggregation throughout the four months of the spawning season.

Significance of Seriola hippos spawning migrations

Migration is best defined in terms of natural selection as it is selection pressures that have formed the observed behaviours that distinguish migration from other kinds of movement (Dingle and Drake 2007). Annual spawning migrations are important life history events that function to connect individuals from large areas to specific sites. It is clear that the aggregation sites near Rottnest Island are important spawning locations for S. hippos inhabiting waters to the south of this area. It is also highly likely that S. hippos

108 individuals investing in long distance migrations to these spawning locations between the months of October and March gain fitness benefits.

Seasonal temperature changes can act as important cues in the timing of migration and for synchronizing reproduction, particularly for ectotherms. Many fish may have temperature dependent gonadal development (Ware and Tanasichuk 1989). The exact spawning season of the circum-global S. dumerili varies depending on the geographical location, however water temperature is a determining factor in gonadal development and the temperatures in which spawning takes place range between 18 and 24°C (Jerez et al.

2006 and reference therein). Seriola hippos aggregations first appear near Rottnest Island during October, at a time when the mean water temperature on the upper continental shelf rises quickly to 18.1°C (cf 16.8°C during September) (Pearce et al. 1999).

The migration of S. hippos from southern waters to the Rottnest Island and west coast region may be linked to the fact that the family Carangidae is of tropical origin. For instance, Australia’s tropical regions hold the greatest carangid diversity with 52 species having distributions primarily north of 23°S whilst only 6 species have primarily temperate distributions (Gunn 1990). The relationship between temperature and gonad development as well as larval development, growth and survival are well documented (see

Rombough 1997). It is therefore likely that S. hippos that are resident in the cooler southern waters migrate up the west coast into more favourable warmer waters for spawning. This idea might also explain the observed minimal movements of S. hippos between the Rottnest Island aggregations and areas to the north. If this is the case, S. hippos individuals resident to warmer northern waters would obviously not have to journey south to find temperatures conducive to spawning.

Further support for the notion that S. hippos gains benefits from, or have a requirement for, spawning in warmer waters may come from the movements of closely related species in South Australia. Preliminary results of a tagging study by Hutson et al.

109 (2007a) suggest that mature S. lalandi return to northern shallow waters of Spencer Gulf seasonally. These tag data together with annual observations of aggregations of large S. lalandi in this region during October have lead researchers to propose that this species migrate into the warmer waters of Spencer Gulf annually for spawning. This possibility has yet to be confirmed, however, Baxter (1960) reported that ocean temperature appears to be a major factor in the seasonal migration of S. lalandi into the Gulf of California.

The reasons why (as opposed to when and where) fish engage in spawning aggregations, often after undertaking long distance migrations, are well documented, and include reducing the level of predation on spawning adults (Thresher 1984) as well as on spawned material (Heyman et al. 2001), increasing mate selection in widely dispersed populations (Claydon 2004) and providing social interaction for sex change determination in hermaphrodites (Shapiro et al. 1993). The fitness benefits, however, behind the temporal (when) and spatial (where) nature of spawning aggregations are contentious and have received less attention. In a review of the major hypotheses Claydon (2004) illustrated that the timing and location of spawning aggregations in fishes are a product of adult predator evasion, decreased egg loss, larval dispersal, larval retention and the pelagic survival of larvae in a patchy environment. Claydon (2004) concluded that due to the overlapping, and often complementary, qualities of these theories it is problematic assigning a single selective force as solely responsible. It is therefore, most likely that a combination of these factors results in the development of spawning times and locations in many aggregating species.

A theme common to most of the hypotheses described by Claydon (2004) is that reproductive strategies employed by many species, particularly broadcast spawners, have been strongly influenced by dominant physical processes, such as ocean currents. This is supported by Hutchings et al. (2002) who reported highly selective reproductive patterns, i.e. spawning times and locations, in many fish species in southern African waters,

110 including clupeids, sciaenids, sparids, and carangids. These authors concluded that natural selection has favoured the synchronization of spawning with ocean current, i.e. adults move up-current for spawning given that prevailing currents transport, then retain, juveniles into productive nursery grounds.

Similar to the African example described above, one of the key processes influencing recruitment success in fish stocks along the lower west and south coast of

Western Australia, particularly those of marine broadcast spawners, is the dispersion or retention of larvae from the spawning grounds by physical oceanographic processes

(Muhling and Beckley 2007). The Leeuwin Current is a dominant annual feature on the west coast of Australia and has been shown to play an integral role in the life history of many marine fishes and invertebrates, including the Western Rock Lobster Panulirus cygnus, Sardines Sardinops sagax, Whitebait Hyperlophus vittatus, Australian Salmon

Arripis trutaceus, and Australian Herring Arripis georgianus (Lenanton et al. 1991,

Ayvazian and Hyndes 1995, Caputi et al. 1996, Fairclough et al. 2000, Gaughan et al.

2001, Muhling et al. 2008a, b).

A review of the dominant oceanic processes of the lower west coast of Western

Australia is offered below to provide background information for two purposes. Firstly, this information will be used in the development and discussion of a hypothesis on the dispersal and subsequent distribution of S. hippos larvae spawned at the Rottnest Island aggregations sites. Secondly, it will be used as a basis to describe what factors may be responsible for the spawning and migratory behaviour observed in S. hippos individuals near Rottnest Island.

Oceanic processes of the lower west coast of Western Australia

The Leeuwin Current is a seasonally varying warm water mass that flows southward along the continental shelf and upper slope, rounding the southwest corner of

111 Australia (at Cape Leeuwin) before flowing eastward across the Great Australian Bight

(Maxwell and Cresswell 1981, Pearce et al. 2006). It is a unique eastern boundary current in that it flows poleward and against the prevailing wind. Flow is strongest during the austral autumn and winter months (March to September), with peak velocities occurring on the lower west coast (32°S) during April and May of each year (Feng et al. 2003). The speed of this current is generally around 0.5 ms-1 (one knot), although speeds of 1 ms-1

(two knots) are common and a maximum speed of 1.7 ms-1 (over three knots) has been recorded (Cresswell 1980). In some years the influence of the Leeuwin Current extends as far eastward as Tasmania (Peter et al. 2005).

During late spring and summer (November to February) mid and inner shelf water movements are more influenced by wind patterns. These waters tend northwards as strong northward wind stresses in the south west of Western Australia slow the Leeuwin Current and, push it offshore (30 to 90 km), and drive the northward flowing Capes Current

(Pearce and Pattiaratchi 1999). The Capes Current is strongest between Cape Leeuwin and Cape Naturaliste after which it broadens along the mid and inner continental shelf as it transports cooler water up the lower west coast and past Rottnest Island (Gersbach et al.

1999). However, because these inner shelf water movements are wind related, meteorological processes cause periodical current reversal and introduce some cross shelf mixing and advection changes during this time (Cresswell et al. 1989, Pearce et al. 2006).

Rottnest Island spawning sites and localised oceanic processes

In a study of larval fish at a series of sites north of Rottnest Island Muhling et al.

(2008a) reported that seasonal assemblages were most closely correlated to water mass, with species distributions reflecting both cross-shelf and along-shore oceanographic processes and events. These authors concluded that the strength and position of the southward flowing Leeuwin Current, and the seasonal, northward flowing Capes Current,

112 drive much of the variability in the marine environment of the region, including the distribution of larval fishes.

Although there has been no long-term detailed study on the physical oceanic conditions in the area where the S. hippos spawning aggregations occur, i.e. 90 to 120 m depth range west of Rottnest Island, there have been several detailed short-term studies. A review of these studies can provide information regarding both the intra and inter-annual variation of water masses affecting the S. hippos spawning locations.

In a reanalysis of the detailed ocean current measurements taken during the Perth

Coastal Waters Study (PCWS, 1991-1994, Lord and Hillman 1995), Pearce et al. (2006), described the outer shelf mid-depth flow over a full year from current meters moored at the same depth in which the S. hippos spawning aggregations are located (110 m). During this time the mean alongshore current component revealed a relatively strong southward direction for most of the year with the strongest currents recorded in March 1992 when the southward component averaged 0.4 ms-1. Although a southward mid-depth current tendency prevailed through the months of October to March (i.e. the S. hippos spawning season), Pearce et al. (2006) described periods of alternating northward flows at this location between October and December (i.e. the early portion of the spawning period).

Details of the prominent near surface water masses encountered over the outer shelf region (100m depth) provided by Muhling et al. (2008a) not only supported the findings of the PCWS but also highlighted the variability in this outer shelf area (their sampling station C, 31° 37’ S 115° 13’E). Changes in the origin of surface water masses within the area of the S. hippos spawning aggregations sites is clearly evident throughout the spawning period. The Capes Current generally prevails during the early months

(October – December), then as the spawning season proceeds, water of Leeuwin Current origin (i.e. southward flow) generally dominates the latter spawning months of January,

February and March. Furthermore, inter-annual variation in prevailing ocean currents and

113 meteorological conditions along the west coast are also common. For instance, during

ENSO (El Nino/Southern Oscillation) years the Leeuwin Current is weaker (Feng et al.

2003), whilst, in other years, quite the opposite occurs and the Leeuwin Current remains strong throughout the summer months (Cresswell and Peterson 1993).

It is highly likely that the eggs and larvae of S. hippos spawning at the Rottnest

Island aggregation sites are subjected to a range of distribution pathways, and thus differing levels of survival, throughout the duration of the protracted spawning season

(October to March) due to those oceanographic variations described above. As a result, spawning at the Rottnest Island sites is likely to produce variable payoffs throughout the spawning season as oceanic dynamics strongly influence the strength of recruitment to different juvenile habitat areas. Disentangling the effects of season, wind stress, temperature, retention and drift trajectories on larval growth and survival is difficult considering that the variables themselves are inter-dependent (Houde 2008). Nonetheless,

S. hippos eggs and larvae produced early in the spawning period are most often subjected to different physical processes than those from later spawning events.

Proposed larval distribution

A schematic detailing potential larval dispersal from S. hippos spawning near

Rottnest Island is presented in Figure 3.9. This figure attempts to show the potential balance of movement of larvae in typical conditions during the spawning period. An important factor to take into account when considering any spawning strategy undertaken by S. hippos is that larval and small juvenile fish may remain within surface waters for at least 60 days after hatching (Chapter 2). Thus, a fish spawned in February may remain under the influence of surface water movements and current flows until April.

114 a) October to December A persistent CC will -31 increase northern Leeuwin Current, LC transport along shelf. Capes Current, CC N Ekman drift may Rottnest Is. Larval Transport periodically force larvae westward. Perth S Spawning Area -32 S

CC flow is sporadic Loss due to weather-related reversals.

-33

Busselton -34 Cape Naturaliste Cape Leeuwin

1 00 m 20 0 m -35 Albany

LC sometimes expands 100 km onto shelf along south coast during this time.

113 114 115 116 117 118 b) January to March

-31 Leeuwin Current, LC Capes Current, CC N Ekman drift may Rottnest Is. Larval Transport periodically force larvae westward. Perth S Spawning Area -32 S

Loss

-33 Occasional CC flow near shore.

Loss Busselton -34 Cape Naturaliste Cape Leeuwin

LC expands onto shelf 1 00 m along south coast. 20 0 m -35 Albany

100 km

113 114 115 116 117 118

Figure 3.9. The proposed dispersal of Seriola hippos larvae from the spawning area near Rottnest Island during, a) the early part, and, b) the latter part of the spawning season.

115 Based on the physical processes that prevail at the S. hippos spawning sites, it is proposed that during the early portion of the spawning period (October - December) the eggs and larvae are most likely to be subjected to current flows and wind patterns that will favour local retention. In addition, the influence of the Capes Current inshore during this time is likely to generate some northward transport (Figure 3.9a) and at the same time the

Leeuwin Current, although weak, will lead to some southward distribution (Figure 3.9a).

Subsequently, as the spawning season proceeds, the net larval drift from the spawning aggregation sites will shift mostly southward as the Capes Current dissipates and a strengthening Leeuwin Current expands into outer and mid shelf waters (Figure 3.9b).

The notion that local conditions are conducive to larval retention during the early months of the S. hippos spawning season is supported by Muhling et al. (2008b) in a study on Sardines Sardinops sagax in an area just north of Rottnest Island. These authors attributed strong egg and larval retention to sporadic flows of the Capes Current due to weather related reversals at that time of year. Muhling et al. (2008b) concluded that a comparatively small proportion of summer spawning production could in fact contribute to the majority of the year’s recruits on the west coast because larvae were held locally in favourable conditions.

Other factors have been identified which may also aid in the transport of S. hippos larvae from the outer shelf spawning sites onto the adjacent upper shelf (i.e. eastward) during this time of year. Characteristic summer wind variation may act to create windows of positive conditions which may lead to onshore larval transport (Cresswell and Peterson

1993, Feng et al. 2003). Furthermore, transport of larvae into nearshore regions adjacent to the south of the spawning area is also likely to occur as tongues of Leeuwin Current water periodically penetrate onto the upper shelf (Figure 3.9a). Pearce et al. (2006), also described a sporadic summer upwelling type process in the lee (northern side) of Rottnest

Island, whereby sub-surface water intrusions mix up onto the continental shelf. This may

116 also facilitate the inshore transportation of S. hippos eggs and larvae. Once on the upper shelf S. hippos larvae are likely to come under the influence of the wind induced Capes

Current which may lead to periodic northward transport depending on its strength and persistence at this time (Gersbach et al. 1999, Muhling and Beckley 2007) (Figure 3.9a).

In addition to being retained locally during the early half of the spawning season, some S. hippos larvae may also be transported considerable distances southward by the

Leeuwin Current which, although generally deemed to be weak during this time, often persists seaward (west) of the 100m isobath (Creswell and Griffin 2004). Larval S. hippos may become entrained in the Leeuwin Current after being driven offshore (i.e. westward) by processes such as Ekman drift (see Gersbach et al. 1999) (Figure 3.9a). These larvae may then be returned shoreward in areas to the south by some of the cross-shelf transport processes described above or as the Leeuwin Current expands onto the upper shelf as it rounds Cape Leeuwin which it sometimes does at this time of year (Creswell and Griffin

2004).

As the S. hippos spawning season proceeds, the Leeuwin Current starts to strengthen (March-April) forming a narrow jet (Feng et al. 2003). The increase in

Leeuwin Current intensity and its consequent expansion into inshore waters as the Capes

Current dissipates at the end of the spawning season is likely to facilitate the dispersal and subsequent recruitment of larvae and juvenile S. hippos into inshore habitats in the southern areas (Figure 3.9b). Furthermore, larvae spawned late in the season (i.e.

February and March) are likely to be picked up by this prevailing current and transported towards the south coast. For instance, between March and April (in most years) the leading edge of the Leeuwin Current progresses eastward after rounding Cape Leeuwin at a rate of no less than 20 km day-1 (Creswell and Peterson 1993). On the assumption that S. hippos larvae are subjected to a current speed of 20 km day-1 over a pre settlement period of 60 days, a potential passive transport distance of 1200 km could be obtained. Thus,

117 larvae spawned in January and February near Rottnest Island might settle into juvenile habitats as far away as Esperance on the south coast.

The Leeuwin Current, however, also periodically produces meanders which carry warm waters over 200km westward into the Indian Ocean (Pearce and Griffiths 1991), a process that would undoubtedly lead to the loss of some S. hippos larvae at certain times throughout the spawning season (Figure 3.9a,b). The high fecundity exhibited by S. hippos, and the batch spawning habits of this species, are likely to help overcome larval losses due to short-term variability. Such loss is a negative aspect of broadcast spawning, an otherwise successful strategy which allows for continual exploration of potential habitats (Hutchings et al. 2002).

The larval distribution proposed above is based on years when typical oceanic processes prevail. During years of uncharacteristic current patterns, for instance, when the

Leeuwin Current is weak, possibly due to El Nino events or when strong wind stresses persist late into the spawning season, larval dispersal during the later months of the spawning season might better be described by Figure 3.9a. Similarly, when the Leeuwin

Current is strong and persists along the west coast throughout the spawning period, for example in years such as 1987 and 2008 (Creswell and Peterson 1993, R. Lenanton,

DoFWA, pers. comm.), larval dispersal during the entire spawning period might be best described by Figure 3.9b.

Possible factors driving the reproductive strategy of Seriola hippos

Numerous marine fish have evolved distinct spawning strategies in order to place their eggs and larvae in favourable locations in the spawning season (Hutchings et al.

2002). Many species that aggregate to spawn annually will adjust the precise location and timing of spawning activities to meet a set of specific environmental conditions that enhance fitness (Claydon 2004). In contrast, the S. hippos spawning aggregations located

118 off the Perth metropolitan coast that form each year do so in a highly consistent manner both temporally and spatially, regardless of highly variable conditions. Thus, as S. hippos does not apparently adjust the location or timing of its spawning, it seems unlikely that S. hippos are synchronizing spawning to a set of specific environmental conditions. This view is also supported by the fact that many S. hippos individuals return to the spawning locations at the same time each year and the fact that many individuals that migrate from distant locations would have no knowledge of spawning site conditions before arrival.

If the location and timing of spawning in S. hippos is not under environmental control then the most likely explanation is that it is genetically inherited. Numerous studies have described that in uncertain environmental condition selection will favour individuals that spread reproductive risk over different years (see Wilbur and Rudolf 2006 and references therein). This bet-hedging reproductive strategy (Slatkin 1974) reduces the impact of environmental variation on reproductive success by increasing the probability that some offspring will encounter conditions favourable for survival (Goodman 1984,

Helfman et al. 1997). As demonstrated in this study, S. hippos, the oldest Seriola species studied so far, commences spawning in its fourth year of life and lives for up to 29 years resulting in up to 25 years of reproductive output. Thus, a S. hippos individual returning at the same time each year (e.g. December) is likely to have an advantage in this variable environment as a fixed strategy will, on average, payoff over the relatively long reproductive lifetime of the individual. In addition to their longevity S. hippos individuals also spawn on multiple occasions for at least one month and at different aggregations sites within the Rottnest Island area, and are thus further bet-hedging within a reproductive season.

Such a strategy can also explain the protracted nature of the spawning aggregations which persist for over four months. For instance, at a location that exhibits both intra- annual and inter-annual environmental variation the payoffs (reproductive success) for

119 spawning at a set time of the year will vary from year to year. Thus, if individuals that arrive and spawn at different times of the season get fitness benefits, and these benefits differ from year to year for each particular time, relatively equal fitnesses will result over time and a protracted spawning season will develop. It appears that the S. hippos population that undertake spawning near Rottnest Island exhibits an evolutionarily stable polymorphic state as defined by Maynard Smith (1982), in that, this population has evolved to a stable equilibrium consisting of individuals with different fixed behaviours.

In other words, over the lifetime of an individual, the reproductive benefit of spawning at a set time each year will be no greater, or less, than that of an individual that spawns at a different period within the spawning season.

Such a fixed reproductive strategy also offers an explanation as to why no apparent correlation exists between the distance of migration and the particular time an individual

S. hippos visits the spawning area. As has been demonstrated the spawning area west of

Rottnest Island is subject to highly variable currents and therefore the offspring of an individual, over its lifetime, are likely to be distributed over much of this species distribution. Thus, if the hypothesis above is true, on reaching maturity an individual will have an innate drive to spawn at a specific time and will move to spawn at that particular time regardless of where it settled as a larva.

It is clear that the Rottnest Island aggregation sites are important spawning locations for S. hippos individuals inhabiting metropolitan and more southerly waters.

Anecdotal evidence collected from an angler recruited for tagging S. hippos in the south- west region offers strong support for the presence of at least one spawning aggregation between capes Leeuwin and Naturaliste during late spring and early summer (D.

Langridge pers. comm.). This fisher reported a large aggregation which persisted at the same location for 3 months in 2005 and again in 2006 with males running ripe and releasing large amounts of milt during the tagging procedure and females containing large

120 roe. Considering that tagging revealed movement of S. hippos from this region to the

Rottnest Island spawning aggregations within the same spawning season, it is plausible that individual fish spawn at multiple locations along the coast during the spawning season. Further support for this notion is offered by the lack of S. hippos with spent gonads at the Rottnest Island aggregation sites in the first half of the spawning season

(Chapter 2), meaning fish were moving in and out of these locations whilst remaining reproductively active. It is plausible that many S. hippos spawning aggregations form along the lower west coast during the spawning season and that individuals may only spend a limited amount of time at one before moving to the next. Such a scenario would

1) further reduce the risk of an individual’s offspring encountering unfavourable conditions, 2) allow larvae to be distributed over a greater geographical area, and 3) provide a greater diversity of mates, all of which increase bet-hedging.

Adult migration and ocean currents

Predictable oceanic currents, which facilitate movement, may be exploited for migration and can possibly influence the evolution of certain migratory aspects (Alerstam et al. 2003). Current systems can be used by fish in strategic ways to optimise migratory economy, thus saving energy. For instance, P. platessa exhibit a 12 hour pattern of vertical movement known as selective tidal stream movement (Metcalfe et al. 2006).

Individuals of this species move up into mid water when the tidal stream flow is in the appropriate migratory direction, whilst remain on the bottom when flow is in the opposite direction. In a similar manner, the Capes Current may provide assistance to S. hippos moving westward around Cape Leeuwin and up the west coast during migration to the spawning grounds in late spring and summer. Likewise, the Leeuwin Current may greatly aid adult S. hippos returning to southern waters.

121 In addition to the potential energy saving by moving with the aid of prevailing currents, S. hippos individuals are also likely to save energy during migration by travelling in groups as indicated by recapture data and the often sudden arrival and departure of fish at fishing grounds. Travelling in schools provides individuals with potential energy savings of up to a 65% (Weihs 1975) and increases endurance during migration due to reductions in hydrodynamic resistance to individuals towards the centre and/or rear of the school (Weihs 1973). For example, study on Horse Mackerel Trachurus mediterraneus, another carangid species, found that the tail beat frequency of trailing fish within a school of was 71-85% the frequency of the leading fish (Zuyev and Belyayev 1970). Similarly,

Herskin and Steffensen (1998) found the tail beat frequency of a Sea Bass Dicentrarchus labrax individual swimming at the rear of a school was up to14% lower than when at the front which corresponded to a 9-23% reduction in the rate of oxygen consumption rate.

Conclusions and future research

The present study has confirmed that S. hippos individuals often undertake long migrations to form transient spawning aggregations near Rottnest Island. The temporal pattern displayed by recaptures also revealed that many S. hippos return to these spawning aggregations at the same time in subsequent years. Upon completion of spawning many S. hippos move from the Rottnest Island spawning aggregations to south-western and southern coastal areas where they are likely to remain resident for much of the non- spawning period. This study is one of a small number that have documented strong spatial and temporal spawning ground fidelity in individuals of a marine broadcast spawner.

Furthermore, evidence suggests that this species is likely to spawn at multiple locations along the lower west coast during the spawning season. The use of electronic data storage tags would greatly aid our understanding of S. hippos spawning behaviour, and potentially provide information on such things as pre- and post-spawning migration routes, timing of

122 migration, period spent at aggregation sites, spawning site fidelity, movement of individuals between spawning aggregations and associated physical parameters such as water temperature.

It is clear that the aggregation sites near Rottnest Island are important spawning locations for southern populations of adult S. hippos. It appears that the S. hippos population that spawn near Rottnest Island has reached an evolutionarily stable polymorphic state, in that, individuals spawning at different times have relatively equal reproductive fitnesses. An extensive search of the literature failed to find any report of the development of this phenomenon, through individuals exhibiting precise spatial and temporal spawning activities, in any other fish species. Future research should be aimed at determining whether such reproductive strategies exist in other Seriola species, or indeed whether it is a common trait of species that form spawning aggregations.

The results of this chapter have led to the development of a hypothesis on larval dispersal. Sampling of eggs and larvae at times and places proposed by this hypothesis would enhance understanding of mechanisms and patterns of dispersal in S. hippos. These results may also be used to develop a better understanding into the ecology/biology of other species in south-western Australia which exhibit similar life history strategies to S. hippos.

Although, tag returns indicate that, spatially discrete ‘southern’ and ‘northern’ adult breeding units may exist along the west coast of Australia, genetic homogeneity is likely to exist due to the exchange of larval and small juvenile S. hippos between these regions. Genetic data are needed to test this hypothesis.

123 Chapter 4

Post release survival of Samson Fish, Seriola hippos Günther 1817, after catch-and-release angling.

4.1 Introduction

Recent fisheries management reforms along the most populated area of the west coast of Western Australia have led to the mixed commercial and recreational demersal finfish fisheries being replaced with recreational only fishing (Anon. 2007). These reforms have followed a trend displayed in many other regions around the world, particularly freshwater environments in temperate regions, where recreational fishers have become the sole user group (Arlinghaus et al. 2002).

In recent times the notion that recreational angling can impact on fish stocks has received increasing attention (Cooke and Cowx 2004, 2006, Coleman et al. 2004,

Arlinghaus and Cooke 2005, Lewin et al. 2006). Recreational fishing participation worldwide is substantial, for instance in 2001 the United States Department of Commerce reported that only 12 % of the entire population had never participated in recreational

124 angling (Cooke and Cowx 2006) and in Australia over one-quarter of the population participate in the recreational and charter fishing sectors (Chapman et al. 2001).

Recreational fishing effort appears to be increasing in most parts of the world with forecasts suggesting that, whilst participation rates in some developed countries are likely to fall as the population ages, the overall number of people fishing will increase (Thunberg

1999, Henry and Lyle 2003). Although a single angler may be considered to have a low catch rate, the cumulative impact of a population can be high. Using data extrapolated from Canadian recreational fishing capture rates, Cooke and Cowx (2004) estimated that the global angling catch could be as high as 47 billion fish annually, of which almost 64%

(over 30 billion) are released.

In addition to fishing closures, both spatial and temporal, two important tools for controlling this increasing fishing pressure are size restrictions and bag limits, whereby fish must be released if under (and in some cases over) a certain size or beyond a certain quantity (Winstanley 1990). These controls are relatively easy to understand and enforce.

A similar concept that (theoretically at least) maximises the enjoyment of fishing whilst ensuring stocks are sustained is catch-and-release angling which is increasingly promoted by recreational fishing bodies such as Recfish Australia, the Australian National

Sportfishing Association, angling clubs and fishing media. Catch-and-release angling is often promoted and applied as a means to cope with high recreational fishing effort and is seen by many as conserving exploited fish populations while maintaining angling use

(Arlinghaus et al. 2007).

Whilst the regulatory measures and voluntary practises listed above can potentially reduce the impact of fishing activities, their success in doing so depends to a large extent on the capacity of the fish to survive capture and steps taken by anglers in capturing/handling the fish that minimise trauma. Although numerous studies have been conducted into the biological aspects of catch-and-release angling, most have typically

125 focused on species or groups of fish that represent the most popular and economically important recreational fisheries (see Arlinghaus et al. 2007 and references therein).

Consequently, the vast majority of this research has been directed towards freshwater species of North America, particularly Largemouth Bass Micropterus salmoides, Atlantic

Salmon Salmo salar and Rainbow Trout Oncorhynchus mykiss. In a recent review and synthesis of 209 studies associated with the biological consequences of catch-and-release angling, Arlinghaus et al. (2007) concluded that only a few generalizations can be broadly applied across catch-and-release fisheries. Although this body of research may be of some use for species for which no data exist, the varied results show that the survival of released fish is generally species-specific due to variations in physiology, morphology and behaviour (Cooke and Suski 2005). The same authors also noted intra-specific differences related to intrinsic (e.g. reproductive cycle) and extrinsic (e.g. water temperature, angling technique) variables. Factors shown to directly affect the survival of a released fish include anatomical hook location (Ayvazian et al. 2002, Broadhurst et al. 2005), fish size

(Taylor and White 1992), terminal fishing gear (such as bait vs lures) (Diggles and Ernst

1997), deep hook removal (Schill 1996) handling time (Ferguson and Tufts 1992), capture depth (St John and Syers 2005) and water temperature (Wilke et al. 1997). It is therefore clear that the effective management of a particular fishery must be guided by research conducted within that fishery.

All aspects of capturing and handling a fish can contribute to stress and/or injury and have a potential to result in mortality. Hooking mortality associated with catch-and- release is generally divided into immediate mortality and delayed mortality (Arlinghaus et al. 2007). Immediate mortality refers to death related directly to capture and up until the time the fish is released. Death which occurs at some time after the released fish has swum away is referred to as delayed mortality, or post-release mortality. In a review of catch-and-release angling studies Bartholomew and Bohnsack (2005) reported that

126 hooking mortality rates across a variety of marine and freshwater species ranged from 0 –

95%. Whilst an angler cannot affect the intrinsic capacity of fish species to withstand catch-and-release the question is one of ensuring fishing practises minimise stress and injury and maximise survivorship. In a paper exploring the future direction of research activities to increase the sustainability of recreational angling, Cooke and Suski (2005) advocate specific catch-and-release research focused towards the development of a suite of specific guidelines for individual species or types of catch-and-release fisheries. These authors also state that research must focus on techniques and equipment utilised by recreational anglers in local fisheries. As such, anglers, researchers, managers and recreational organisations are increasingly realising that they have a responsibility to:

1) Develop specific protocols for recreational fisheries that minimise the trauma to fish,

2) Educate fishers on why such protocols should be followed, and

3) Ensure that these protocols are followed.

The overall aim of this chapter was to investigate the effect of catch-and-release within the Samson Fish Seriola hippos sportfishery in Western Australia (see Chapter 1) with a view, if appropriate, to develop protocols aimed at maximising the survival of released fish. Two approaches were undertaken towards this objective. As detailed in

Chapter 3, the first approach was to undertake a large scale collaborative research project titled ‘Samson Science’ with the intention to engage the recreational fishing community in the research of this popular sportfish species. Samson Science was centred around the tag and release of S. hippos from the Rottnest Island spawning aggregations over two seasons, the summers of 2004/05 and 2005/06.

The above collaborative approach was undertaken in the belief that the best way to deal with the issues of stress/mortality is through research, involvement of anglers, education, and development of practical, well investigated fishing protocols. Thus, the

127 plan was not only to develop protocols for best fishing and fish handling practices, but to ensure that fishers were made well aware of these practices and voluntarily adopt them.

User groups and stakeholders are more likely to accept research findings when they can personally relate to the subject area or issue and when they can participate in problem identification, scientific activities, or meetings to interpret data. Furthermore, a high level of community engagement is often cost-effective and when recreational fishers are treated as partners, with the results communicated in a timely manner using language they understand, strong community support for necessary management reforms is likely (Leslie et al. 2004). Such an approach also leads to greater stakeholder satisfaction and adoption of research outcomes, and thus increased community stewardship of the resource. Indeed, many of the reforms witnessed in recreational fisheries in other parts of the world by

Arlinghaus et al. (2007), with respect to how fish are treated during catch-and-release, have arisen as a result of anglers, scientists and managers working together. In addition to the benefits of angler participation described above, the Samson Science component of this study aimed to collect data to describe angler behaviour, investigate effects of different tackle and determine immediate mortality.

A second approach to study the effects of catch-and-release angling on this species was to examine any physical trauma and physiological stress that may contribute to post- release mortality. Observations of released S. hippos, which generally swam strongly towards the bottom, were indicative that, at least in the short term, post-release survivorship is likely to be high. However, such an assumption can not be drawn solely from these observations as Muoneke and Childress (1994) showed that most mortality caused by catch-and-release occurs sometime after release. An experimental enclosure was therefore used to monitor survival of S. hippos for up to 31 hours after capture in order to determine if there was any delayed mortality.

128 Two major factors associated with the delayed mortality of captured fish are physical injuries and the cumulative effects of physiological stress (Olla et al. 1998,

Cooke and Sneddon 2006). Physical trauma is manifested as external and internal tissue or organ damage that is generally caused by hooking, on-board handling and barotrauma

(Skomal 2007, Rummer and Bennett 2005). Of most concern for anglers within the S. hippos sportfishery was injuries related to decompression (see Chapter 1) because of the depth of water in which the aggregations are found (> 90m), a capture depth that would almost certainly be lethal to local demersal species such as the West Australian Dhufish

Glaucosoma hebraicum (St John and Syers 2005).

The effects of major physiological stress, in particular, the cumulative effects of numerous stressors such as exhaustive exercise, air exposure, handling and temperature, can also induce mortality. Thus, a second point of concern for the welfare of released S. hippos was the stress induced during the capture on these powerful fighting fish and their ability to recover from exhaustion. The ability to respond to stress is a vitally important mechanism of all organisms. Although the word stress often has negative connotations, the stress response allows fish to avoid or cope with challenges to homeostasis (Barton 2002).

The physiological response to stress is considered an adaptive mechanism that preserves homeostasis or allows homeostatic recovery. Physiological response mechanisms may, however, be compromised and become maladaptive or dysfunctional if the intensity of a stressor, or stressors, is overly severe or enduring (Wendelaar Bonga 1997). Physiological responses to stress are classified into primary, secondary and tertiary responses depending on the mechanisms involved.

The primary stress response in fish involves the initial neuroendocrine reaction.

This response includes the rapid release of catecholamines (such as adrenaline) from the chromaffin cells and corticosteroid hormones (principally cortisol) from the interregnal cells, each of which are located in the head kidneys, into the blood stream (Wendelaar

129 Bonga 1997; Mommsen et al. 1999). An elevation of plasma cortisol is the most widely used indicator of stress in fish and is typically used to measure physiological disturbance, and thus stress (Wendelaar Bonga 1997). Secondary stress responses include changes in blood and tissue ion and metabolite levels related to the effects of the released hormones as well as to physiological adjustments caused by processes such as respiration, metabolism, cardiac output and cellular responses (Mommsen et al. 1999). Changes in plasma lactate and glucose are typically used to describe such responses in fish after a stressful event (Wendelaar Bonga 1997). Tertiary stress responses are detrimental long- term effects which refer to the fish’s overall function such as growth, immune responses, reproduction and survival.

It has been well documented that capture and handling of fish during angling causes acute physiological responses (see Pankhurst and Sharples 1992, Pankhurst and

Dedual 1994, Wang et al. 1994, Barton 2002, Arlinghaus et al. 2007). These studies cover a wide range of teleost species and have established characteristic physiological responses in fish subjected to catch-and-release by measuring the changes in various haematological parameters after physical exhaustion. Generally the majority of physiological changes induced by a catch-and-release event have returned to resting levels within 2-24 hours, depending on the variable examined.

The objective of this part of the study was to gain insight into some of the physical and physiological consequences of angling on S. hippos and to characterise and quantify stress reactions and recovery times of this species released after angling. The main focus was to determine if physical traumas and homeostatic disruptions, due to capture depth, high anaerobic activity and exhaustive exercise, which may impede normal physiological and behavioural functions, will reduce survivorship. Thus, the aim was to generate estimates of mortality from catch-and-release fishing which can be used by management agencies.

130 There is a recognised need to ensure that the sportfishery meets certain standards aimed at minimising stress to the animal and maximising survivorship. It is accepted that recreational angling is increasing and that people will continue to go out and catch fish, be it for food or fun. Thus, if catch-and-release is shown to have major detrimental effects on the S. hippos spawning aggregations then this information can be used to make appropriate management decisions. If, however, there appears to be negligible effects on individuals or the spawning aggregations then the results of this chapter, together with personal observations and discussions with anglers, will be used to establish best handling protocols tailored specifically for the Western Australian S. hippos sportfishery. These protocols will also included statements about the impacts of fishing on S. hippos together with catch-and release guideline generalities, so anglers can make informed choices regarding fish welfare.

4.2 Materials and Methods

Tag data

Data collected during the tagging component of the Samson Science project

(detailed in Chapter 3) were used to investigate fishing methods and release techniques used within the Rottnest Island sportfishery.

Enclosure trials

A large open topped cylindrical floating enclosure or ‘sock’ measuring 15m deep x

2m diameter was used to assess the short and longer term post-release survival of S. hippos (Figure 4.1). This apparatus was designed by the Queensland Department of

Primary Industries and Fisheries to investigate the short term post release survival of tropical demersal species. The Samson Science project was able to borrow the enclosure and secure vessel time for 10 days in both early March 2006 and in February 2007.

131 Two research vessels of the Department of Fisheries Western Australia, the RV

Naturaliste and RV Snipe, and the charter boat North Star 2 were used during the enclosure trials. The RV Naturaliste, a steel research trawler, was used to deploy and retrieve the enclosure and monitor fish behaviour. The smaller RV Snipe and North Star 2 were used to catch and transport fish to be placed into the enclosure. For most of the trials the Naturaliste drifted with the sock attached alongside, or, if the water depth was less than 60 m the Naturaliste was anchored (Figure 4.1b).

a) b) c)

Figure 4.1. The sock enclosure used to hold and monitor Seriola hippos during the survival trials, a) being deployed, b) tethered to the RV Naturaliste, c) containing fish with the live feed video camera in the upper portion.

To keep surface interval times (i.e. the time taken to transport fish and release them into the enclosure after capture) as low as possible the Naturaliste deployed the sock upwind of the fishing area while the second vessel was used to catch and transport fish (up to 3 fish at a time) to the sock. The surface interval time was therefore dependent on the drift speed of the RV Naturaliste through the fishing area and the time taken to catch and transport fish to the enclosure. The surface interval time was mostly kept to less than 10 minutes. Up to 11 S. hippos were placed into the sock at any one time. The length of time each fish was held in the enclosure varied because fish were placed into the enclosure at

132 separate intervals (due to only transporting up to 3 fish at a time) yet all fish in the enclosure were retrieved at the same time. For instance, on one occasion it took over 4 hours to put 10 individuals into the enclosure due to factors such as fish catchability, weather and fish transport time. Trial times for individual fish, therefore, lasted for between 2 and 31 hours as the Naturaliste drifted with the wind and current or was at anchor.

All fish were captured using a metal jig attached to a heavy (70 to 90 kg) monofilament leader with a 37 kg main line, a method typical of the sportfishery. Once a fish was landed on the RV Snipe the time was recorded and the fish were measured, tagged and placed into a 200 litre rectangular transport tank with lid, filled with fresh seawater. Each fish was tagged in a different position along the body (right/left, and 5 positions along the body in relation to the second dorsal fin) so each was distinguishable when monitored by video in the enclosure. If/when a second or third fish was hooked, fresh seawater was allowed to flow into the tank and the lid closed until that/those fish were landed. The time was recorded when each fish was released into the enclosure after transportation. A summary of the number of S. hippos used during the short term post release survival trials is presented in Table 4.1.

Table 4.1. Summary of the fish held in the enclosure for the 2006 and 2007 delayed mortality trials. N.B. The 2007 trials included one male fish captured in 55 m depth.

Number of Fish Hours in Capture Year Male Female Total enclosure depths 2006 21 34 55 2 – 31 34 – 195 2007 17 4 21 2 – 21 80 – 113 Totals 40 36 76

During the trials fish health and behaviour were monitored continually, during daylight hours, with a live feed video camera (Figure 4.1c). Each individual could be distinguished on the video by the location of the tag, and every 30 minutes, each was

133 assigned a ‘health’ score. Observed swimming behaviour categories (1. Swimming upright; 2. Lays on side occasionally; 3. Often on side; 4. Always on side) and fish colour

(1 Uniform bronze; 2. barred/blotched) were used to generate a health score for individual fish within the enclosure. Health was assigned as: 5 - bronze and swimming upright; 4 - barred/blotched and swimming upright; 3 - barred/blotched and on side occasionally or often; 2 - barred/blotched and always on side. A score of 1 was assigned to any fish that died. On the completion of each trial the fish were euthanized for biological sampling and assessment of barotrauma related injuries.

As the trials were conducted towards the end of the spawning season in both years, problems were encountered catching fish from the aggregations normally targeted by sportfishers (in depths of ca 110 m). Consistent catches of S. hippos during the time of the

2006 trials were found on a nearby wreck (HMAS Derwent) located in waters of depth

195 m, due West of Rottnest Island (32° 03.458’S, 115° 12.298’E), and also shallow waters north of Rottnest Island (ca 34 m).

Blood physiology

Blood samples were taken from a subsample of 35 fish at the completion of the enclosure trials, in order to investigate the stress response and recovery after angling.

Winching equipment aboard the RV Naturaliste allowed for fast retrieval of the enclosure and blood samples were taken from recovering fish within 3 minutes of initial enclosure disturbance. As fish were placed into the enclosure at different times, plasma cortisol and lactate levels were determined at a range of times post release, thus allowing for estimates of recovery times. Additional blood samples where collected from fish immediately after capture by angling. A 2 to 4 ml blood sample was taken from the caudal vasculature of each fish with a 22 gauge hypodermic needle. Blood was transferred immediately into heparinized vials and placed on ice. Blood plasma was separated in all samples by

134 centrifugation within 10 minutes of collection, then frozen and stored at -80°C to await analysis.

Plasma cortisol was measured by a competitive chemiluminescent assay using a

Bayer Advia Centaur® analyser. Plasma lactate was determined by the amperometric method using an ABL800 FLEX blood gas analyser (Radiometer Medical ApS,

Copenhagen, Denmark).

4.3 Results

Tagging and engagement of fishers

Information detailing specific fishing methods and handling techniques was recorded for 5464 S. hippos tagged during the intensive tagging events undertaken west of

Rottnest Island (Chapter 3). The percentage of these fish caught and released by fishers recording this information is summarised in Table 4.2. Of these fish 3356 (61.4 %), and

2108 (38.6 %) were caught and released by fishing charter boats and small boat fishers, respectively. Most (94%) fish were caught using artificial metal jigs whilst the remainder were captured using baited hooks. Most anglers used fishing line with breaking strains over 30 kg. Gut hooking, which included the stomach, oesophagus and gills, was rare with most fish hooked in the jaw area of the mouth (97.8%). A small number of fish (1.8%) were classed as foul hooked which mostly included hooks lodged in the operculum and in the head behind the eyes or in the body near the pectoral fin.

The majority of fish (~ 95%) were successfully released by use of the spearing method, whereby the fish is thrust head first into the water as the body weight is supported by one hand behind the head with the other around the caudal peduncle. Whilst most fish subjected to this release method swam down when released on the first attempt (97.7%),

2.3 % were only released successfully with 2 or more attempts after floating following the initial release effort. The majority of the remaining fish, which were deemed by fishers to

135 require other modes of recovery or which floated after a failed spearing attempt(s), were released by the use of a release weight (see section 4.4) (~ 3%), whilst 0.8% were revived by flushing the gill with a deck hose and 0.8% were revived by towing. Overall 0.6 % of

S. hippos died due to capture.

Table 4.2. Summary of the details collected by fishers involved in tagging Seriola hippos at the sportfishing grounds west of Rottnest Island. The fishing gear, position of hook, lift and revive methods, and release condition as a percentages of the 5464 Samson fish caught and released during the 2004/5 and 2005/6 seasons. N.B. gut refers to stomach, oesophagus and gills.

Method/technique Percentage Fishing method Bait 5.8 Jig 94.2

Line class (kg) < 15 2.2 15 – 30 32.9 30 + 64.9

Hook position Mouth 97.8 Gut 0.4 Other 1.8

Lift method Net 1.4 Leader 64.7 Leader and tail 25.0 Other 1.9

Revive method Speared 95.7 Towed 0.8 Deck hose 0.8 Release weight 2.7

Release condition Healthy 94.7 Floated/revived 4.7 Died 0.6

The mean FL of fish that floated after the initial release attempt was significantly higher than the overall mean fork length of fish tagged during each season (2004/05, t =

6.69, P < 0.001; 2005/06, t = 10.05, P < 0.001) (Table 4.3). The overall number of fish that floated was higher (5.7%) during the first season of tagging compared to the 2005/06 season (3.8%). This difference may, therefore, be attributed to changes in angler

136 behaviour and increased fish handling awareness through project communications and

extensions as there were no changes to the tagging program design between seasons. For

example, the use of a release weight increased notably during the second season of

tagging, particularly during the first attempted release, which may account of the lower

number of fish that floated during that season (Table 4.3).

Table 4.3. The total numbers and mean lengths of Seriola hippos that were tagged and released in the 2004/06 and 2005/06 seasons and the numbers and size of those that floated and/or were released with the use of a release weight. Numbers in parentheses are the percentage of overall fish tagged during the specified season.

Use of Release weight Total Mean FL Number Mean FL On 1st Overall Season tagged (cm) floated (cm) floaters floaters attempt SS1 2005 2427 107.0 137 (5.7) 113.5 31 5 36 (1.5) SS2 2005/06 3037 107.1 115 (3.8) 117.9 47 58 105 (3.5)

Enclosure trials - general observations

Over the 10 day period in May 2006, 55 fish were used in the short term post

release survival trials. During March 2007 a further 21 fish were captured for use in the

enclosure trials. Trial times for individual fish ranged from 2 to 31 hours. Of the fish

captured for these trials, 40 were males ranging in length from 520 – 1275 mm FL with an

average of 919 mm FL (±29 mm S.E.). Females (n = 36) ranged from 575 – 1120 mm FL

with an average FL of 878 mm (±22 mm S.E.). All fish used in the trials were hooked in

the jaw and had no major hooking injuries.

Seriola hippos released into the enclosure spent the majority of time towards the

bottom portion where they grouped together generally facing into the current, or

alternatively, when there was little current, most fish generally swam around in a circle

formation. Larger fish (>1200 mm TL, n = 4), however, appeared to have difficulty

turning around in the enclosure (diameter of 2 m) and were often observed to circle up and

down between the bottom and mid portions when there was little current flow through the

137 netting. Once settled in the enclosure most fish showed little sign of stress, even when, during one trial, the sock was continually circled by a large Hammerhead Shark (Sphyrna sp.) for four hours.

Enclosure trials - survival

The transport time between catching the fish and releasing them into the enclosure

(i.e. surface interval time) was found to be a critical factor for the survival of fish captured at the deep water Derwent site (195 m). The delayed mortality of these fish greatly increased with surface interval times of over 20 minutes (Figure 4.2). Fish from this site that took longer than 20 minutes to transport were therefore not included in delayed mortality or health observation analysis (n = 8). Moreover, these high surface interval times, whereby these fish caught in very deep water and transported in shallow containers, are not indicative of the sportfishery where such fish would be released relatively quickly and presumably return to the depth of capture.

6 100 80 - 113 m depth 195 m depth 11 7 17 6 5 80

60

40 Survival (%) Survival

20

5 3 0 < 10 10 - 20 20 - 30 > 30 Surface interval time (min)

Figure 4.2. Survival of Seriola hippos caught from 80-113m and 195 m depth subjected to different surface interval times.

138 Overall six of the 64 fish (9.4 %) placed in the enclosure during the trials died, representing and overall survival rate of 90.6%. Survivorship of fish caught at 34 m depth and held in the enclosure for between 21 and 25 hours (n = 12) was 100%. Ninety three percent of S. hippos captured at depth ranging from 80 to 113 m (n = 28) survived their duration in the enclosure, which ranged from 2 – 22 hours for different individuals (Table

4.4, Figure 4.3). The two fish that died after capture from this depth range were both males caught at 110 m depth. One of these fish became entangled in the netting of the enclosure during the trial, the other died within 4 hours of release into the enclosure. Enclosure trials involving fish caught from a depth of 195 m (n = 24) lasted for between two and 30 hours in which 83% of fish survived (Table 4.4). Fish that died after capture from this depth comprised three females and a male, and in each case, mortality occurred within the first five hours (Figure 4.4).

The lengths of fish that survived (n = 58) ranged from 520 to 1182 mm FL and had a mean of 914 mm FL (± 21 mm S.E.), whilst the length range of the fish that died ranged from 640 to 1275 mm FL with a mean of 897 mm FL (± 87 mm S.E.). There was no significant difference between the lengths of the fish that survived and died throughout these trials (t = 0.248, P = 0.805) or between the survival of males and females (Pearson

χ2= 0.021, P = 0.883).

Table 4.4. Mortality of Seriola hippos from three depths of capture.

Capture Number Number Mortality Depth Caught Died (%) 34 12 0 0 80-113 28 2 7 195 24 4 17 Total 64 6 9.4

All, but one, fish caught from 34 m depth had a health score of 5 when first released into the enclosure. The remaining fish had a health score of 4 upon release and

139 recovered fully within the first hour (i.e. health score 5). All fish maintained a health score of 5 throughout the trial (21- 25 hours).

The average health score assigned to fish that survived capture from 80 to 113 m depth (n = 24) increased markedly within the first hour. All fish within this group that survived appeared to show normal colour and behaviour within 5.5 hours of being released

(Figure 4.3). There was no significant difference between survival of fish caught from depths of 80-113 m and 195 m (Pearson χ2= 1.148, P = 0.284).

5.0

4.5

4.0 Mean health score Mean 3.5

3.0 0123456789 Time (hours)

Figure 4.3. Mean health score of Seriola hippos that survived capture from 80 – 113 m depth. Note two fish died after capture from this depth range.

The mean health score of fish that survived capture at 195 m depth reached a score of 5 (i.e. all fish appear healthy) within 4 hours post release (Figure 4.4). Fish that survived capture from this depth had a higher health score upon release into the enclosure

(i.e. time = 0) than those fish that died at sometime during the trials.

140 5

4

3 Mean health score Mean 2

1 02468 1416 Time (hours)

Figure 4.4. Mean health score of Seriola hippos that survived (dashed line) and died (solid line) after capture from 195 m depth. Note all mortality occurred with 5 hours.

Physical effects of capture

Overall six fish (8.1% of fish examined internally) captured from depth ranging from 34 to 195m showed signs of internal haemorrhaging or clotting. Haemorrhage was observed in the posterior region of the body cavity and mostly attributed to damaged blood vessels (Figure 4.5). Of these fish, 5 survived for between 4 and 24 hours before being euthanized for internal examination at the end of the respective trial by which time three fish were categorised with a health score of 5 (i.e. healthy) and two had a health score of 4

(i.e. barred/blotched and swimming upright). The only other fish with signs of internal bleeding died after 5 hours in the enclosure. Internal bleeding/clotting was noted in 8.3%

(1), 14.2% (4) and 2.9% (1) of fish captured from depths of 34m, 80-113m and 195m, respectively. Gross displacement of viscera and compaction injuries to organs due to swim bladder over-inflation were not observed in any of the fish examined.

141

Figure 4.5. Photograph of an internal blood clot in a Seriola hippos captured from 195 m depth.

Six fish examined internally had ruptured swim bladders. In each case this consisted of small perforations or lateral tears of less than 10mm in diameter or length

(Table 4.4, Figure 4.6). No fish captured at 34 m depth showed any signs of swim bladder damage, whereas 3.6% (1) and 15.6% (5) of fish captured from 110m and 195m depth, respectively, had ruptured swim bladders. The survival rate of fish that had ruptured swim bladders and were included in the post release survival trials (n = 5) was

80%.

Table 4.5. Percentage of Seriola hippos that had intact and ruptured swim bladders caught from depths of 34, 81-113 and 195 metres. N.B. not all fish were examined post- mortem; * includes one fish captured at 55 m.

Swim Bladder (%) Depth (m) Intact Ruptured

34* 100 (12) 0

80-113 96.4 (27) 3.6 (1)

195 82.4 (28) 15.6 (5)

142

Figure 4.6. Photograph of a ruptured swim bladder in a Seriola hippos captured from 110 m depth.

Blood physiology

Blood plasma samples were taken from 35 S. hippos in March 2006 and February

2007 during the enclosure trials. These fish ranged in fork length from 630 to 1182 mm with a mean of 908 (± 26.4) mm. Large increases in both mean plasma cortisol and plasma lactate levels were evident after capture. Each of these indices increased relatively quickly and peaked soon after the capture event (4 hrs for plasma cortisol and 2 hrs for plasma lactate). It must be noted that, due to the logistical constraints of sampling, each of these indices may have in fact peaked before these times yet no measurements were taken at those time periods. This was followed by a gradual decrease over a longer period of time (Figures 4.7 and 4.8).

143 300 4

250 5

1

) 200 -1

3 5 150 3 2 3

Cortisol (ng ml (ng Cortisol 100

2 3 4 50

0 0 5 10 15 20 25 30 35 Recovery Time (h)

Figure 4.7. Mean (± S.E.) plasma cortisol levels in Seriola hippos for each recovery period. Sample sizes for each time interval are shown.

35

1 4 30

25 ) 5 -1

20

15 LactateL (mmol 10

4 5 5 2 5 3 3 3

0 0 5 10 15 20 25 30 Recovery Time (h)

Figure 4.8. Mean (± 1 S.E.) plasma lactate levels in Seriola hippos for each recovery period. Sample sizes for each time interval are shown.

Mean plasma cortisol levels ranged from 40 ng ml-1 immediately after capture

(time 0), and peaked at 259 ng ml-1 4 hours post capture before declining to 60 ng ml-1 31

144 hours after the stressful event (Figure 4.7). Although plasma cortisol levels decreased markedly after 5 hours, these levels remained higher at 31 hours than those recorded at capture.

The mean plasma lactate level in S. hippos was 5.4 mmol L-1 upon capture by angling and peaked 2 hours after capture at 31 mmol L-1. This level then declined below the concentration at capture after 17 hours to 1.9 mmol L-1 and remained below the level at capture for the duration of the trial (i.e. 30 hours) (Figure 4.8).

4.4 Discussion

Mortality

Tagging data collected during this study demonstrated that capture related death before release, i.e. immediate mortality, associated with angling S. hippos at depths typical of the sportfishery was less than 1 % (Chapter 3). However, the delayed mortality of S. hippos captured at these depths was much higher (7 %). Thus, the findings of this study are in agreement with that of other studies concerning catch-and-release mortality (see for example Muoneke and Childress 1994), in that most observed death is delayed and occurs sometime after release. The delayed mortality figure determined for S. hippos however, should be viewed as indicative only, because logistical difficulties associated with studying a large species in a location well offshore did not allow for the use of control fish or a large sample size. The use of experimental control fish under these circumstances would be extremely difficult to achieve, however, since no mortality was observed in confined fish captured at 34 m depth, mortality associated with confinement is likely to have been negligible.

The total hooking mortality of S. hippos subjected to catch-and-release angling within the Rottnest Island sportfishery is approximately 8 % (i.e. immediate mortality plus delayed mortality). Tagging undertaken as part of the Samson Science project revealed

145 that a massively concentrated fishing effort, far beyond normal fisher behaviour, could catch up to 2000 fish per month, or 8000 fish per season within this fishery (Chapter 3).

Based on a conservative mortality estimate of 10% together with a high estimated total catch-and-release of 8000, fishers targeting S. hippos could cause the unintended death of up to 800 fish during a season. Thus, considering a worse case scenario of 1000 S. hippos deaths per year together with a very conservative estimate totalling 67 200 fish (Chapter 2 estimate for only one aggregation), total mortality attributable to catch-and-release angling represents, at most, only 1.7% of the S. hippos population spawning near Rottnest Island.

However, considering the large number of S. hippos moving through this area during the season, for instance, 67 200 to 150 000 individuals estimated seasonally at the Outer Patch aggregation, one of only four targeted aggregations within an area where 14 are known to exist (A. Bevan, Shikari Charters, pers. comm.), the level of mortality associated with the catch-and-release of this species is, indeed, going to be less than the above estimate and likely to have little effect on the S. hippos population.

The results of this study whereby mortality of S. hippos increased to 17 % for fish captured in a water depth of 195 m, reflect similar trends to other such studies showing that mortality increases with capture depth due to barotrauma (Feathers and Knable 1983,

Rummer and Bennet 2005, St John and Syers 2005, Stewart 2008). However, S. hippos are much less susceptible to depth induced mortality than other species in which barotrauma has been observed. St John and Syers (2005), for instance, found that mortality of G. hebraicum increased from 21% at depths of 0-14m to 86% at depths of 45-

59m. Similarly Feathers and Knable (1983) established that the mortality of M. salmoides increased from 25% for fish caught in 9m to over 50% for fish in 27m. Post release mortality rates over 20%, as found in the two studies described above, are considered to be unacceptably high for catch-and-release fisheries and deserving of management attention

(Muoneke and Childress 1994). Even though the rates of mortality in S. hippos captured

146 in 195m approach this figure (i.e. 17%) it should be noted that these waters are far deeper than those fished by anglers in the sportfishery.

The results of this study demonstrated that S. hippos captured from deep water (i.e.

195m) had much greater mortality when subjected to long surface intervals (i.e. > 20 minutes held in well aerated water in an onboard tank) than fish that were promptly released into the enclosure. This large increase in mortality is likely due to longer confinement periods during transportation. Longer confinement periods may have resulted in increases in confinement stress and the effects of decompression. For instance, fish held on the surface (i.e. at low ambient pressure) for long periods after ascent from deep water would have greater exposure to the effects of supersaturated dissolved gases, such as nitrogen, and the formation of gas bubbles as the solubility of gasses in the blood and tissues is decreased (Henry’s Law) (St John 2003). Thus, observations of higher survival in S. hippos subjected to shorter surface times might best be explained by the mitigation of barotrauma, such as gas embolisms, as a result of prompt release into the enclosure and allowing fish to return to depth, albeit only 15 meters.

Physiological stress response in Seriola hippos

This study demonstrated that cortisol elevations in S. hippos, in response to an acute stressor (i.e. capture and handling), are within the 30 to 300 ng ml-1 range characteristic of other fishes subjected to stress events (Barton 2002). Although plasma cortisol values can vary markedly depending on species, stock, developmental stage, nutritional state and environmental factors (Barton 2002), S. hippos reacted to catch-and- release with a typical teleost plasma cortisol response (Mommsen et al. 1999). For instance, S. hippos demonstrated the general overall trend whereby the concentration of cortisol in the blood increased significantly within 4 hours following the stressor. The elevated cortisol levels in S. hippos then displayed a gradual decline and exhibited a

147 similar recovery profile to other fish species following acute stress, such as Barramundi

Lates calcarifer (de Lestang et al. 2004), Pink Snapper Pagrus auratus (Pankhurst and

Sharples 1992), Carp Cyprinus carpio (Pottinger 1998) and two species of coral trout

Plectropomus leopardus and P. maculatus (Frisch and Anderson 2005).

Although plasma cortisol levels decreased markedly after 5 hours, these levels remained higher at 31 hours than those recorded at capture. A similar occurrence was described by de Lestang et al. (2004) who investigated the effects of catch-and-release angling on barramundi and suggested that elevated cortisol levels after 24 hours post release were possibly due to additional stress associated with the retention of fish in holding nets during the recovery period. Furthermore, research on salmonids has suggested that active swimming may enhance post release recovery and that confinement in an enclosure can lead to artificially long recovery times (Milligan et al. 2000, Farrell et al. 2001). It is therefore likely that S. hippos released under normal fishing conditions would experience faster recovery than the fish that had their natural swimming ability impeded by containment during the present study.

Elevated plasma lactate levels, which are a secondary stress response, occur as a consequence of respiration under anaerobic conditions in which glycogen supplies are depleted and lactate accumulates in the white muscle fibres (Milligan and Girard 1993).

Levels of plasma lactate are, thus, not used as a measure of stress, as such, but instead are used to determine the time required for recovery from the respiratory effects of exhaustive exercise or a strenuous event (Pankhurst and Dedual 1994; Pottinger 1998). As was the case with plasma cortisol, recovering levels of elevated plasma lactate in S. hippos were consistent in duration with the general trend documented for other species for which complete recovery occurred within 24 hours post release, such as L. calcarifer (de Lestang et al. 2004), C. carpio (Pottinger 1998) and O. mykiss (Pankhurst and Dedual 1994).

148 The results of this study established that the capture and release of S. hippos with rod and line evoked a neuroendocrine stress response, and that the exhaustive exercise associated with this capture resulted in metabolic disturbances. Although baseline plasma cortisol and lactate levels of undisturbed S. hippos where not recorded in this study, each of these levels decreased to below, or near to, capture levels during recovery.

Furthermore, since all mortality observed during the enclosure trials was attributable to barotrauma related injuries such as internal haemorrhaging, post release mortality in S. hippos does not appear to be caused by physiological dysfunction associated with the stress of capture. The results demonstrated that each of these primary and secondary stress responses in S. hippos subjected to catch-and-release angling are within physiological tolerance limits. Difficulty in the determination of baseline pre-stress physiological indicators in large fish species was highlighted by Skomal (2007) since acquiring blood samples requires some degree of capture and handling. In the case of S. hippos this problem is further compounded by the depths at which fish are located. Perhaps one solution would be to spear (i.e. with spear gun) free swimming individuals that occasionally follow hooked fish to the surface then quickly draw blood.

Handling Protocols

Based on the data presented above, observations made by researchers and discussions with anglers the following protocols have been tailored specifically for the Western

Australian Samson Fish sportfishery. These best handling practises also include statements about the impacts of fishing on Samson Fish so anglers can make informed choices.

1) Almost all lure (metal jig) (Figure 4.9a) caught fish were hooked in the hinge of the

jaw and very few were bleeding. Similarly, other studies have shown that the use of

149 lures results mostly in minor superficial injuries and near zero hooking mortality

(Diggles and Ernst 1997, van der Walt et al. 2005). The standard pattern for rigging a metal jig is with a single short shanked “assist hook” attached by a heavy synthetic cord. Whilst these hooks are not true circle hooks as defined by Cooke and Suski,

2004, (i.e. a hook in which the point faces directly at the hook shank rather than being parallel to the shank), they do have circle hook characteristics in that the shank is shortened, the entire hook is rounded, and although the point does not face perpendicular to the shank it does start to bend towards it. Current studies from elsewhere in the world indicate that the use of circle hooks with no or only a minimum offset (i.e. the plane of the hook point is parallel or close to the plane of the shank rather than the plane being offset by as much as 15°) (Figure 4.9b and 4.9c) result in less deeply hooked fish and less trauma (see Cooke and Suski, 2004 for a review of hook damage by different style hooks). Circle hooks have been shown to have lower capture efficiency than other conventional hook types in some species, however, this is invariably a result of the anglers striking at bites as they would when using normal J- shaped hooks (Cooke and Suski 2004). As they are designed to roll over and catch in the hinge of the jaw as the fish takes a bait, turns and swims away, striking has the effect of pulling a circle hook out of the fish’s mouth. However, even without a change in fishing practices this type of hook is still very effective in the capture of

Samson Fish, as even though several strikes may not result in hook-ups at most sportfishing sites, hook-ups will inevitably occur due to the abundance and voracious striking nature of this species. Thus, jigs with single barbless assist hooks are the preferred method of capture, whilst bait fishermen should use circle style hooks with minimum offset. Apart from being easier to remove, barbless hooks have also been shown to reduce tissue damage (Meka 2004). If barbless hooks are unavailable anglers can use pliers to crush or break the barb on barbed hooks.

150

Figure 4.9 a) A metal jig used to catch Samson Fish. Note the single assist hook, with barb crushed, attached to the front of the lure, b) and c) a non-offset circle hook (left) and an offset J-hook right.

2) Analysis of tag sheets and personal observations on the water suggests that the longer

it takes to bring the fish to the boat, the more likely difficulties will occur when

releasing the fish, i.e. the released fish will not swim strongly back to the school and

may float on the surface. The appropriate line classes, i.e. 24kg and above, should be

used, lighter line classes than this should not be used to gain points in competitions or

for ‘bragging rights’. Evidence from other studies shows that long capture times often

result in increases in lethal and sub-lethal effects (Cooke and Sneddon, 2006).

3) In their natural environment Samson Fish are supported by water. When possible this

large species should be left in the water alongside the boat for de-hooking and release.

Barbless hooks can simply be removed and the fish orientated into the ‘spear’ position

and released (see protocol 5). If fish have to be brought onboard, do not gaff them, do

not use grips to lift them by the lower jaw, and do not lift them just by the tail or gills.

Preferably support each fish by both the head and tail, or if necessary use a knotless

151 landing net. A length to weight conversion table is provided in the Samson Fish

handling protocols booklet so anglers can estimate weight from fish length to reduce

the need to weigh an unsupported fish.

4) The sock trials and tagging data sheets show that the longer fish are out of the water

the greater the chances of mortality. It is also likely that sub-lethal trauma increases

with time out of water. These findings are in agreement with several published studies

that consider reducing the time out of water is a critical factor in maximising the

survival of released fish (Ferguson and Tufts 1992, Cooke and Sneddon 2006). To

minimise fish exposure time the following should be adhered to: a) unhook fish in the

water if possible, b) use barbless hooks or crush the barb on barbed hooks to make

unhooking easy and fast, c) if deep hooked, cut the line as other studies have shown

that this significantly reduces mortality and the fish is likely to shed the hook on its

own (Diggles and Ernst 1997, van der Walt et al. 2005), d) if fish are removed from

the water for dehooking and/or tagging, have all necessary equipment at hand, work

swiftly and ensure that fish are placed on a wet, cool surface, e) if fish are removed for

photographs, ensure that the camera is ready and everyone knows where the

photographer and subject will stand/sit before removal from the water.

5) The most effective way of returning Samson fish is to ‘spear’ them back into the

water. This procedure entails supporting the weight of the fish with one hand just

behind the head and under the gut, and the other around the caudal peduncle (the

narrow wrist section anterior to the tail). The fish is then speared head first into the

water, in the vast majority of cases the fish will swim strongly back to the school

below (underwater video shows just how quickly fish recover when this method is

used). In contrast fish that are dehooked and then gently released tend to swim

152 downwards for a short distance before resurfacing and floating off. Spearing fish back

into the water also seems to be effective for other species such as Pink Snapper Pagrus

auratus and Silver Trevally Pseudocaranx dentex. However, care should be taken not

to slam the fish back into the water with excessive force.

6) Data gathered during tagged events revealed that large individuals have a greater

tendency to float upon release. Anglers should therefore exercise additional care and

control during an attempted spear release or use a release weight if deemed necessary

(see Protocol 7).

7) If a fish returns to the surface and floats, lines should be immediately brought in,

fishing ceased, and the fish followed and retrieved as quickly as possible. On retrieval

the health of the fish should be assessed, if the fish appears relatively healthy the

spearing method should be reattempted. Alternatively, if the fish is deemed to be tired

or has an inflated abdomen it should be released with the aid of a release weight. The

release weight is a large lead sinker to which is attached a large barbless hook that has

a swivel and clip on its shank and a crimp on its bend, the latter stopping the swivel

from coming free (Figure 4.10). The weight is attached to a hand-line or rod and reel

(preferably all ready set up and specifically used for this purpose). The hook is then

placed in the jaw of the fish and the fish released, the weight quickly takes the fish to

deeper water, compressing the swim bladder and gas in sinuses as it descends. Once

the fish has reached approximately 40 metres a series of sharp tugs on the line frees the

fish, it is worth noting that one can often feel the fish start to swim strongly and release

itself well before this depth is attained.

153 In the case of fish that the angler deems to be less healthy or completely exhausted

(little or no movement and shallow and rapid movement of the gill covers) or if the release weight needs to be made ready, the fish should be swum by the side of the boat or brought onto a wet, cool boarding platform and have a deck hose placed in its mouth, this procedure ensures that well oxygenated water is passed over the fishes gills. Fish revived using a deck hose should have their eyes covered with a wet hand or towel to prevent sunlight damaging their retina (Brill et al. 2008). Once the fish shows signs of revival

(deep slow movement of the gills and strong tail beats) it should then be attached to a release weight and released. Although the low number of recaptures precludes statistical validation of the effectiveness of the release weight, the fact that tagged fish that anglers considered in very poor condition and released using this method have been recaptured after more than a year at liberty suggests that the method can only benefit the welfare of the animal.

a) b)

Figure 4.10. a) A release weight used for the release of Samson fish to aid descent. Note: the hook is barb-less for release at depth with minimal effort, b) A West Australian Dhufish Glaucosoma hebraicum suffering barotrauma (note bulging eyes) being recompressed and released with the aid of a release weight. Photograph courtesy G. Lilley.

154 8) Various shark species, including Shortfin Mako Isurus oxyrinchus occasionally appear

and predate on hooked fish at the aggregation sites west of Rottnest Island. In

addition, shark predation at times can be particularly problematic at the northern

aggregation site off Hillarys where schools of whaler sharks Carcharhinus spp. often

prey on fish that are hooked or upon their release. Immediate mortality caused by

acute injuries or predation when fish are hooked at this location can often approach

100%, to the point where fishers are unable to land a single fish. If Samson Fish are

constantly being taken by sharks fishing should cease and anglers should move to

other fishing grounds.

9) Samson fish are much less susceptible to barotrauma injuries than many other local

species. However, barotrauma induced mortalities do occur in the water depths of the

Sportfishing sites west of Rottnest Island (i.e. 80-113m), and thus, anglers should

target this species in waters as shallow as possible.

In the case of S. hippos this study has perhaps been fortunate in working on an animal that is apparently particularly robust in that it rarely suffers obvious signs of barotrauma, i.e. extruded stomach, bulging eyes or gross internal damage. This is probably a result of certain anatomical novelties found in this species (see Chapter 5). The majority of the protocols developed are those that would apply to most species, however, it appears that the spearing technique is not promoted or used anywhere else. This technique also seems to work well with P. auratus and P. dentex and is therefore likely to work with a wide range of other catch-and-release species promoted through popular media.

Currently many authorities around the world advise anglers to ‘vent’ or ‘fizz’ fish that are showing obvious signs of barotrauma. This procedure involves puncturing the swim bladder through the side of the fish with a hypodermic needle (Kerr 2001). This method,

155 however, has received considerable debate with many studies on post-release survival producing contrasting results (e.g. Wilson and Burns 1996, St John and Syers 2005,

Rummer and Bennet 2006). Although this procedure may increase survival rates in the hands of experienced researchers using sterile techniques (McLeay et al. 2002), it is less likely to have as good a result when performed by the majority of anglers who are unlikely to use sterile techniques and hypodermic needles, particularly on large fish such as S. hippos. Another point of consideration is that the venting procedure in S. hippos and other large fish is slow and takes a few minutes to complete due to the large volume of excess gas. This situation may therefore increase barotrauma due to a longer surface time creating greater exposure to the effects of excess dissolved blood and tissue gases before the fish can be returned to capture depth. Furthermore, venting is obviously a major invasive procedure, that at a minimum causes damage to the body wall and swim bladder, may lead to infection, and will possibly affect the ability for fine positional control thereby reducing the ability to feed and avoid predators (Kerr 2001). For these reasons it is recommended that in cases of obvious barotrauma a release weight be used rather than venting.

The protocols and related discussion above address all but one of the generalised catch-and-release guidelines that would be applicable to the S. hippos sportfishery as outlined by Cooke and Suski (2005). These include reducing the duration of the angling event, minimising air exposure and the use of barbless hooks and artificial lures. The remaining guideline relevant to S. hippos yet to be addressed is the avoidance of angling a species immediately prior to or during the reproductive period. Chapter 2 clearly demonstrated that the S. hippos sportfishery west of Rottnest Island indeed targets fish aggregated for spawning. Catch-and-release angling during the spawning period has potentially negative fitness consequences since the reproductive period is essential for generating offspring. Acute and chronic stress in reproductively active fish caused by

156 capture, confinement and handling, can induce endocrine alterations and/or changes in behaviours that depress fitness (Pankhurst and Dedual, 1994, Clearwater and Pankhurst,

1997, Morgan et al. 1999). For example, O. mykiss subjected to acute stress during spawning showed delayed ovulation, reduced egg size, reduced sperm counts and a decreased in offspring survival (Campbell et al., 1992). Similarly, Morgan et al. 1999 showed that, although Atlantic Cod Gadus morhua subjected to a chronic stressor were able to spawn successfully, fish initiated fewer courtships, often performed altered courtship sequences and produced abnormal larvae more frequently. Determining the fitness consequences of catch-and-release angling is logistically difficult to undertake, particularly in marine broadcast spawners, and it is possibly for this reason it has received little research attention. Several freshwater studies, however, have documented clear fitness impacts of catch-and-release, particularly on species that exhibit parental care. For example, catch-and-release angling has been shown to decrease fitness in M. salmoides, a species that displays parental care in the form of a nesting strategy. Angling of M. salmoides was shown to reduce reproductive success due to increased predation of unprotected offspring whilst the male was removed from the nest (Philipp et al. 1997) and the impairment of parenting abilities if the male eventually returned (Cooke et al. 2000).

In terms of marine species, Lowerre-Barbieri et al. (2003) investigated the effects of catch-and-release on the reproductive output of actively spawning Common Snook

Centropomus undecimalis captured at a spawning aggregation. Based on acoustic telemetry and histological evidence from tag recaptures these authors concluded that the stress of catch-and-release angling did not cause fish to immediately leave the aggregation or interrupt spawning. Interestingly, even fish implanted with ultrasonic tags continued to spawn.

Any potential impacts on the fitness of S. hippos associated with catch-and-release within the Rottnest Island sportfishery are likely to be reduced by a high seasonal

157 abundance of this species and the fact that the locations of many spawning aggregations are not known by most fishers. The worst case scenario, based on estimated seasonal catches made earlier, is that catch-and-release angling will stop the input of up to 10000 fish per year at the aggregation sites. In terms of management, fisheries authorities and stakeholders need to be made aware of this potential impact so informed management decisions and choices can be made, respectively.

Another area that may be worthy of investigation is whether there are other factors which may affect the spawning success and fitness of S. hippos at the spawning aggregations. For instance, does the presence of naval exercises or large numbers of whaler sharks in close proximity to the spawning sites result in noticeable change in the behaviour of S. hippos as anecdotal evidence suggests?

Angler uptake and feedback

During this study there was a noticeable increase in awareness and voluntary uptake of these suggested best practices. This is likely to be a result of the ongoing dialogue with recreational anglers via clinics, training and information sessions, and web chat sites. For example, without telling anglers to change their fish landing and releasing methods it was found that in 2005/06 less fish were reported as ‘floaters’ than in 2004/05

(115 cf 137) despite an increase in the number caught-and-released (3037 cf 2427). This also coincided with anglers using the release weight 105 times compared with only 36 times in 2004/05. In addition, when a released floating fish was seen it was not uncommon for another boat that was closer to the fish than the boat which had released it to stop fishing, retrieve the fish and then use a pre-rigged release weight to successfully return the fish to the depths. It was also noticeable that the vast majority of anglers targeting the S. hippos aggregations were well aware of the need to use appropriate tackle to catch these powerful fish. Inputs from fishing tackle retailers also suggests that anglers

158 not directly involved in the project but targeting these fish, and other species, are buying suitable equipment and release weights.

Implications for research and management of recreational fisheries

It is unlikely that the numbers of recreational fishers will decrease in the immediate future (Thunberg 1999). With the continuing promotion of catch-and-release in popular fishing media (e.g. http://www.westernangler.com.au) and by government agencies (de Lastang et al. 2004), it is highly likely that the proportion of anglers practicing catch-and-release will increase. It is therefore imperative that researchers and anglers are aware of their responsibilities to develop, promote and use fishing and fish handling practices which minimise trauma and thereby maximise the survival of fish they release. As different species react to the trauma of fishing in quite different ways, researchers must focus on developing protocols that are species specific (Cooke and Suski,

2005).

The development of such protocols for individual species would, in most cases, be beyond the scope of most research groups without the aid of recreational fishers and/or charter boat operators. The current project has demonstrated that by using intensive training and education sessions, where techniques and the scientific method were clearly explained in lay terms, recreational anglers can provide robust scientific data that can be used to develop such protocols. A further benefit that comes from a proper collaboration between researchers and anglers is that both parties gain a better appreciation of and increased respect for the fish they are catching. Anglers also gain a great deal of satisfaction knowing that successful collaborations with researchers is more likely to lead to the long term sustainability of their sport and are therefore more likely to participate in other projects.

159 Once protocols and associated management decisions have been developed it is incumbent on researchers and managers to explain these, and the science behind them, in appropriate, clear and concise language. If the reasons for changes to practices are not made easily accessible then it is unlikely that they will be accepted and many anglers will just carry on using inappropriate techniques. For example, Lucy and Davy (2000) state that the potential benefits of management efforts, such as bag and size limits, can be negated when anglers do not understand reasons behind, or support the underlying principles for, regulations. Thus, direct engagement and extension towards stakeholders must not be overlooked as an effective management tool.

160 Chapter 5

A novel swim bladder allows rapid ascent in kingfishes (Seriola spp)

5.1 Introduction

With the exception of bottom dwelling species all fishes face the dilemma of maintaining their position in the water column, rather than sinking to the bottom.

Different groups of fishes have evolved different methods of attaining neutral or near neutral buoyancy. For example, extant pelagic sharks use a combination of the generation of lift through the shape of their head, shape and angle of their body and pectoral fins, and a tail fin that has an upper lobe that is larger than the lower lobe; a reduction in weight, due to the skeleton being composed of cartilage rather than bone; and the presence of low density compounds in their tissues and organs, such as lipids, in particular squalene in their especially large livers (Corner et al. 1968, Helfman et al. 1997, Phleger 1998, Moyle and Cech 2004). Whilst a hydrodynamic body, reduction in skeletal and muscle mass, and incorporation of low density compounds reduces the energy requirements of maintaining their position in the water column, such adaptations can carry certain penalties. For example, the reliance on hydrodynamic aids means forwards propulsion must be maintained, reductions in skeletal and muscle tissue may reduce the activity of an animal,

161 whilst lipids are expensive to produce (Moyle and Cech 2004). Buoyancy in these species can also vary as lipids are metabolized during periods of starvation (Pelster 1998).

Perhaps the most elegant solution that fishes have evolved in order to maintain their position in the water column is found in the bony fishes. In this group a flexible- walled, gas-filled chamber, the swim bladder, is used to provide buoyancy. As a fish rises in the water column pressure decreases and the swim bladder increases in volume making the fish more buoyant so gases are removed, conversely as a fish descends the bladder becomes smaller and gases must be added to the bladder to stop the fish from sinking.

The swim bladder has two significant advantages over other methods of buoyancy control.

Firstly, gas can be added to, or removed from, the bladder to provide truly neutral buoyancy. Secondly, now that the form of the fins and body are no longer constrained by a requirement to provide lift they can be modified to provide increased efficiency in propulsion (tail fin), fine positional control such as hovering (pectoral fins) and the development of a diversity of body forms suited to particular strategies. Thus, the evolution of the swim bladder and it’s freeing of fin and body shape from providing lift, along with changes to the structure of the feeding apparatus, are probably the main reasons why bony fishes are the largest and most diverse of all vertebrate groups (Gregory 1933,

Helfman et al. 1997, Berenbrink et al. 2005).

Two main types of swim bladder are recognised depending on how gases enter and leave them, which is related to the morphology of the bladder and the systematic position of the fish. Lobe-finned fishes (Sarcopterygii), such as lung fishes, and basal ray-finned fishes (), such as bichirs, gars, bony tongues, eels, , carps and minnows, catfishes, and pikes, salmonids and smelts, have a swim bladder that is directly connected to the oesophagus by a duct throughout life (physostomous swim bladder).

Gases enter (in the most part) and leave the swim bladder through this duct with fishes gulping air at the surface or “burping”air out depending on whether they want to descend

162 or ascend. Although some physostomes can remove and add gases to the swim bladder through diffusion via glands similar to those described below, the majority appear to have to gulp air at the surface if they need to replenish gas in the swim bladder. For this reason the majority of physostomes are found in shallow waters or relatively close to the surface

(Moyle and Cech 2004).

In contrast to the physostomous condition, the more derived spiny ray-finned fishes (Actinopterygii: Acanthopterygii), such as mullets, dories, sticklebacks, scorpionfishes, perches and flatfishes have a swim bladder that is separated (at least in juveniles and adults) from the oesophagus. In this condition gases enter the swim bladder via a bundle of anteroventral capillaries, the rete mirable, and associated gas gland. The gas gland produces lactate and hydrogen ions that diffuse into the incoming capillaries thereby reducing the pH of the blood, this reduction in pH lowers the ability of haemoglobin to bind to oxygen and other gases (Bohr and Root effects) which is thus given up to the plasma (Root 1931, Ball et al. 1955). Once the partial pressure of gases in the blood becomes higher than that in the swim bladder they diffuse across the gas gland and into the bladder. Movement of gases into the bladder is further aided by the fact that the addition of lactate reduces the solubility of gases in the blood (salting out) and thus more oxygen, and also carbon dioxide and nitrogen diffuse into the swim bladder

(Alexander 1966, Copeland 1969). The counter current system of the capillaries in the rete ensures that the high concentrations of lactate, hydrogen ions, oxygen, carbon dioxide and nitrogen are maintained within this region (Wittenberg et al. 1964). The swim bladder is in general lightly vascularised and lined with sheets of guanine crystals that reduce leakage of gases from the bladder (Pelster 1998). Gases are removed from the swim bladder by diffusion back into the blood stream via capillaries on the dorsal wall of the swim bladder, these may be densely packed in a region called the oval or more diffuse if

163 the bladder has a separate compartment controlled by a muscular sphincter (Moyle and

Cech 2004).

The anatomy of the physoclistous swim bladder allows very precise positioning and also reduces the reliance on filling at the surface, which is presumably why physoclists are more diverse in deep water than are physostomes. However, the presence of a sealed flexible-walled gas chamber within the body cavity imposes certain constraints

(Strand et al. 2005) as changes in the volume of, and thus buoyancy provided by, the swim bladder are related to two things. Firstly, what type of changes in depth occur, and secondly, how fast can gases can be secreted and released from the bladder. Take for example, a fish at the surface where there is 1 atmosphere of pressure, as it descends to 10 meters the pressure increases by 1 atmosphere and the swim bladder will halve in volume, as the fish continues to descend there will be further increases of 1 atmosphere of pressure per 10 meters and a proportional decrease in the size of the swim bladder such that if no gases are secreted into the bladder it will be one tenth of its initial volume at 90 meters.

Thus, a fish that descends faster than gases can be secreted into the swim bladder will become negatively buoyant and will have to use additional energy and hydrodynamic aids to maintain its position. In contrast, a fish at 90 meters that moves to the surface its swim bladder would, in theory, have to undergo a ten-fold increase in volume and would become more and more positively buoyant as it ascends. Thus, if a physoclist were to undertake a rapid ascent over a large pressure differential there is likely to be a point when the rapid increase in positive buoyancy cannot be overcome and the fish will float to the surface. Therefore, individual physoclists tend to remain within fairly tight depth limits

(Arnold and Greer-Walker 1992) or undergo relatively slow daily vertical migrations that may result in the individual being negatively buoyant for much of the time (Harden-Jones and Scholes 1985, Parker et al. 2008). For example, Strand et al. (2005) modelled that a 2 kg Atlantic Cod Gadus morhua undertaking a one hour migration from depths of 200 m to

164 between 150 and 50 m would have to have a swim bladder that provided between -40 and

0% of neutral buoyancy. Similarly, in telemetry studies, Parker et al. (2008) concluded that Black Rockfish Sebastes melanops undertaking diel vertical migrations maintained neutral buoyancy at shallower depths but were negatively buoyant at greater depths.

From the above it is clear that physoclists do not normally undergo rapid vertical movements. The effects of sudden decreases in ambient pressure are well known by anglers when fish are pulled to the surface during capture (Figure 5.1). This rapid decompression often results in swim bladder overexpansion that in its most severe forms leads the bursting of the swim bladder, eversion of the stomach, intestinal prolapse, damage to the liver and other organs, and loss of equilibrium, other signs of barotraumas are exophthalmia (bulging eyes), and internal and external haemorhaging, (Brueswitz et al. 1993; Parrish and Moffitt 1993; Wilson and Burns 1996; Rummer and Bennett 2005).

Figure 5.1. Photograph of a Mulloway Argyrosomas japonicus captured from 110m water depth displaying the barotrauma symptoms of stomach eversion and exophthalmia.

165 During the current study the release of gas by Samson Fish Seriola hippos on ascent when hooked was noticed. Indeed, almost all S. hippos observed during capture at the spawning aggregation sites (90 to 110 m depth) released large quantities of gas from the opercula region, particularly during the last 10 to 20 m before reaching the surface (Figure

5.2). As S. hippos are a carangid fish within the Acanthopterygii and thus physoclistous, it could be assumed that the release of this gas was associated with a rupture of the swim bladder as commonly occurs with other angled physoclistous species (Rummer and

Bennet 2005). However, free swimming S. hippos individuals which followed hooked fish to the surface were also observed to release gas during ascent, whilst conversations with SCUBA divers and rock lobster fisherman revealed that they have also witnessed this species undertaking rapid vertical movements and the release of gas. As the release of gas on ascent by this physoclist appears to occur under “natural” conditions it is unlikely that it results from swim bladder rupture, however this can not be discounted. In the larval stages of many physoclist fishes the swim bladder is initially connected to the oesophagus, after the swim bladder is inflated for the first time by gulping air at the surface the connection is subsequently lost (McCune and Carlson 2004). Thus, a second explanation is that this species has retained a connection between the swim bladder and the oesophagus. A third explanation is that S. hippos have developed another mechanism to release excess gas from the swim bladder during rapid vertical movements. The aim of this Chapter, therefore, was to describe how this gas is released and thereby determine which of these hypotheses is correct.

166

Figure 5.2. Photograph of gas release of from under the operculum of a Seriola hippos captured in water of 110m depth. Photograph taken at 5m depth courtesy of Garry Lilley.

5.2 Materials and Methods

Seventy four S. hippos that were captured from depths ranging from 34 to 195 m were collected for external and internal examination. An additional five fish from a water depth of 110m were frozen at sea soon after capture in order to keep the swim bladder in situ for examination by x-ray imaging. X-ray images of these frozen fish were taken at the radiology laboratories in the School of Veterinary Biology and Biomedical Science at

Murdoch University. In the laboratory careful dissection was used to determine the morphology of the swim bladder. In addition, a two-part silicone (with a curing catalyst), which was initially flowable and then cured to a gel, was introduced into the swim bladder under slight pressure via a syringe to determine the route of escaping gas.

Two other Seriola species, S. lalandi and S. dumerili, and the following carangids were also dissected for comparative purposes: Carangoides chrysophrys, Onion Trevally Carangoides caeruleopinnatus, Blue Spot Trevally Caranx bucculentus, Yellowtail Scad Trachurus novaezelandiae, Black-banded Amberjack

Seriolina nigrofasciata and Talang Queenfish Scomberoides commersonnianus.

167 5.3 Results

Initial dissections of S. hippos following capture from depth showed that typically the swim bladder was fully inflated but rarely was it ruptured (Table 4.5, Chapter 4). In addition, signs of haemorrhaging of blood vessels were noted in 8.1 % of all fish examined (see Chapter 4). External examinations subsequently confirmed the presence of a single ovate or circular hole generally 6-10 mm in diameter in an area of soft tissue in the dorsal region underneath each operculum (Figure 5.3). Generally a single hole was present on both the left and right sides of an individual, in some specimens however, a discernible opening was only visible on one side (Table 5.1). Such holes were found in

86, 100, and 97% of fish captured from 34, 80-113 and 195 m depth respectively. The percentage of fish with an obvious hole under both opercula increased with depth (Table

5.1). As demonstrated in Chapter 4 no fish captured at 34 m depth showed any signs of swim bladder damage, whereas 3.6% and 15.6% of fish captured from 110m and 195m depth, respectively, had ruptured swim bladders.

Figure 5.3. Hole under the operculum where gas is released from the swim bladder of Seriola hippos during ascent.

168

Table 5.1. Percentage of Seriola hippos that had obvious signs of opercular venting caught from depths of 34, 81-113 and 195 metres. N.B. * includes one fish captured at 55 m; Number of observations in parentheses.

Vented (%) Depth (m) None LHS RHS Both

34* 14 (2) 50 (7) 28 (4) 7 (1)

80-113 0 21.4 (6) 25 (7) 53.6 (15)

195 3.1 (1) 21.9 (7) 6.25(2) 68.8 (22)

Dissection and x-rays demonstrated that the swim bladder of S. hippos, like most other physoclists, is located just below the vertebral column in the peritoneal cavity

(Figure 5.4). Similarly, the gas gland and underlying rete mirabile are located in the anterio-ventral region of the swim bladder in this species. The swim bladder of S. hippos internally possesses a dorso-posterior excretory chamber that is separated from the secretary chamber by a partition with an aperture controlled by a muscular sphincter connecting the two cavities.

Internal examination of the dorsal chamber of the swim bladder revealed a membranous opening in the posterior region of its roof (Figure 5.4). This opening leads into a membranous tube that runs forward between the dorsal surface of the swim bladder and the spine (Figures 5.4, 5.5). Towards the anterior of the swim bladder, at approximately precaudal vertebra 4 and 5, the flattened membranous tube splits laterally

(left and right) around the spine before exiting under each operculum (Figures 5.4, 5.6).

169

Figure 5.4. Top left photograph; the opening in the dorsal surface of the swim bladder that allows the movement of expanding gases through to the hole underneath each operculum (top right photo). The x-ray image (of a Seriola hippos of 1090 mm TL) showing the path taken by expanding gases through the opening and along membranous flat tube towards the opercula. The dashed line denotes split of tube as gas is directed to each operculum.

10mm

Figure 5.5. Silicone cast of the membranous tube that connected the dorsal surface of the swim bladder to an exit hole in the opercular region of a Seriola hippos of 395 mm TL. N.B. the silicone flowed to one exit hole only after the tube split laterally; cast viewed from above.

Dissection of S. lalandi and S. dumerili revealed a homologous structure in each of these species. The other carangids, however, did not possess this structure rather they exhibited the normal physoclistous condition.

170

a) Swim bladder

Stomach Pneumatic duct Intestines

Dorsal Aorta Ovale To b) heart

Constrictor muscles Gas Gland

Rete mirabile with parallel vessels dissected apart c) Swim bladder vent

Highly vascularised ovale

Exit hole Gas Gland Constrictor muscles

Operculum

Rete mirabile

Figure 5.5. Swim bladders: (a) a physostomous fish with swim bladder connected to the digestive system by the pneumatic duct; (b) vascular connections of a physoclistous swim bladder which is disconnected from the digestive system; (c) the physoclistous swim bladder of Seriola hippos which is connected externally by the swim bladder vent. N.B. (a) and (b) from Pough et al. 1996.

171 Discussion

In this Chapter it has been demonstrated that Seriola hippos and at least two other members of the nine species of Seriola have a unique swim bladder not found in the other carangids examined, including one of their closest relatives Seriolina nigrofasciata

(Smith-Vaniz 1984).

The release of gas by S. hippos during rapid ascent has been witnessed in relatively natural circumstances. Seriola hippos have been observed to vent excess swim bladder gases on ascent at high speed following rock lobster pots being winched to the surface by a commercial vessel in the Western Rock Lobster Fishery (C. Radford pers comm.).

Seriola hippos are regularly observed chasing pots to the surface during retrieval then remaining in wait for discarded bait. The pot winches used in this fishery generally retrieve lobster pots at a rate of approximately 90 m per minute (J. Oreb, Hamilton

Engineering, pers comm.). Recreational SCUBA divers have also reported this species venting gas on ascent. One diver, whilst descending on a dive near Busselton, observed several individuals in a school of 15 to 20 S. hippos release gas after the school had ascended off the bottom to meet him mid-water, presumably out of curiosity (D. Lane pers comm.). Underwater gas release was also often observed during the current project at the deep water (110m) aggregation sites when non-hooked fish followed hooked fish to the surface, these fish had presumably ascended from considerable depth.

As the anatomical structure described above, vents swim bladder gases to the exterior, and as angler referred to the release of gas as venting, it is here after referred to as the swim bladder vent. The swim bladder vent acts in such a way as to provide a pressure release valve for the swim bladder during ascent by providing a path of least resistance for expanding gases when the internal pressure of the swim bladder is increased. It remains unknown as to whether or not the membranous opening of the swim bladder vent is under active control or operates passively. Either way, it is apparent that when a ‘critical’

172 internal swim bladder pressure is reached during an ascent expanding gases exit via the flattened membranous tube and vent into the water. It is likely that venting would continue whilst the internal pressure exceeds this ‘critical pressure’ threshold during ascent. Then, when the fish levels off, or turns to descend, internal swim bladder pressure would drop below ‘critical pressure’, venting would cease and the fish would retain a fully functioning undamaged swim bladder.

This unique swim bladder enables S. hippos to undertake rapid vertical movements that might otherwise result in severe barotrauma injuries. The ability of a physoclist to expel excess swim bladder gases whilst retaining full swim bladder function would offer the following advantages.

1. Prey capture

Each of the Seriola species in which this structure has been identified, S. dumerili,

S. hippos and S. lalandi, can be considered pelagic/semi-pelagic species (Thompson et al.1999, Moran et al. 2007) whose diet include a high proportion of clupeids and squid

(Baxter 1960, Andaloro and Pipitone 1997, Chapter 2). These prey can undertake rapid vertical movements as the former are physostomes whilst the latter do not have swim bladders. Thus, an individual with this structure would be better able to take advantage of these types of prey than an individual with the usual physoclistous type of swim bladder.

2. Predator avoidance

This type of swim bladder would also offer a considerable advantage in the avoidance of predators in a pelagic environment, particularly if a predator is not subjected to the constraints of a physoclistous swim bladder. For example, Shortfin Mako Sharks

Isurus oxyrinchus, whaler sharks Charcarhnius spp. and Blue Sharks Prionace glauca are

173 often seen at the S. hippos spawning aggregation sites near Rottnest Island (A. Rowland, pers. obs).

3. Spawning behaviour

The spawning behaviour of S. hippos, which is known to occur in deep water (i.e.

90 to 120 m), is likely to involve spawning rushes which comprise high speed vertical movements, an attribute typical of aggregating pelagic spawners (Claydon 2004). Such action is also likely to facilitate in the release of gametes as the expanding swim bladder presses onto gonads. Thus, the ability to undertake extended vertical movements would allow for the release of more sperm or eggs. Furthermore, it is likely to provide energy saving benefits because individuals may have to undertake fewer ascents followed by positively buoyant descents.

The presence of the swim bladder vent would almost certainly explain observations by recreational fisherman which suggest that these species do not suffer the typical external symptoms of barotrauma. The swim bladder vent provides S. hippos with an ability to overcome all of the major physical injuries that have been documented to be related to an over inflated swim bladder associated with rapid ascent during angling, including organ displacement, bruising and compression of organs, everted stomach/esophagus and exopthalmia (Brueswitz et al. 1993; Parrish and Moffitt 1993;

Wilson and Burns 1996; Rummer and Bennett 2005). This study demonstrated that the incidence of apparent venting holes occurring under both opercula increased with depth of capture. There appears to be limitations as to the rate at which expanding swim bladder gases can escape via the swim bladder vent during rapid ascent. This rate is sometimes exceeded during line capture when S. hippos individuals are pulled to the surface and is reflected in the higher incidence of swim bladder rupture at greater capture depths.

Although rapid capture is advocated (protocol 3 in Chapter 4), it may be advisable to slow

174 retrieval over the last 10 to 20 m in order to allow rapidly expanding swim bladder gases to escape. As water clarity at many of the deep water fishing sites allows fish to be seen within 10 to 20m of the surface and, as many anglers use braided line that has colour coded depth marks, reducing the retrieval rate as fish approach these depths should be relatively easy. This advice has been included in the protocol booklet provided to anglers.

In this chapter a novel feature of the physoclystous swim bladder that allows rapid ascent in three Seriola species has been described. Further work is required to determine whether or not the swim bladder vent opens passively or is under muscular control, the extent of its use under natural conditions and whether it is found in other Seriola species and their relatives.

175 Chapter 6

Summary and General Conclusions

This thesis had two overriding aims. The first was to describe the biology of

Samson Fish Seriola hippos and therefore extend the knowledge and understanding of the genus Seriola. The second was to uses these data to develop strategies to better manage the fishery and, if appropriate, develop catch-and-release protocols for the S. hippos sportfishery. The underlying hypotheses tested in the individual studies conducted within this thesis are included in the following discussion where appropriate.

6.1 The biology of Seriola hippos

Reliable biological information on age structure, growth, mortality rates, length and age at first maturity, spawning period and fecundity is essential for developing fisheries management plans for a species. This study represents the first comprehensive study of the age, growth and reproductive biology of S. hippos, one of the nine species in the genus Seriola and an important commercial and recreational species. Three of the larger members of the genus, S. dumerili, S. lalandi and S. quinqueradiata, have been subjected to numerous studies due to their economic importance.

Similar to S. dumerili and S. lalandi, the annuli in sectioned otoliths of S. hippos represent annual features and are, therefore, appropriate for determining age and growth parameters of this species. The formation of the opaque zone in the otoliths of S. hippos occurs during late spring/summer (November - January), which is in contrast to many other species with similar temperate distributions in Western Australia such as the West

Australian Dhufish Glaucosoma hebraicum, Mulloway Argyrosomas japonicus and Silver

Trevally Psudocaranx dentex in which the annual opaque zone becomes delineated with a new translucent zone during spring (Hesp et al. 2002, Farmer et al. 2005). Instead opaque

176 zone formation in S. hippos is correlated to the summer spawning period when reproductive development of the gonads, limited feeding and large-scale migration occur.

Seriola hippos displays similar growth trajectories to other Seriola species that have received research attention (Baxter 1960, Manooch and Potts 1997a, Gillanders et al.

1999a, Thompson et al. 1999, Wells and Rooker 2004). The growth of juvenile S. hippos is rapid with this species reaching minimum legal length for retention of 600mm TL within the second year of life. Fast growth continues during the first 5 years of life until around the size of sexual maturity (~830 mm FL). After this the growth rates of each sex differ, with females growing faster and reaching a larger size at age than males. Thus, by

10, 15 and 20 years of age, the predicted fork lengths (and weights) for females were 1088

(17 kg), 1221 (24 kg) and 1311 mm (30 kg), respectively, compared with 1035 (15 kg),

1124 (19 kg) and 1167 mm (21 kg), respectively for males. This observed size difference is most likely as a consequence of the group spawning strategy of this species. A larger size in female S. hippos is favoured because it increases fecundity, whereas male size is most likely driven by male-male competition for females and egg fertilization (Parker

1992). Despite these differences, male and female S. hippos attain similar maximum ages, i.e. 28 (1280 mm FL) and 29 years (1470 mm FL), respectively. The maximum age of 29 years recorded in this study indicates that individuals of this species may live considerably longer than other large members of the genus. For instance S. dumerili has been recorded to reach 17 years (1365 mm FL) (Manooch and Potts 1997b), whilst S. lalandi has been documented to attain an age of 21 years (136 cm FL) (Stewart et al. 2004). Such higher longevity might be explained by the fact than many S. hippos individuals inhabit cooler higher latitude waters (Heibo et al. 2002).

Seriola hippos, a serial spawner with indeterminate fecundity, has a protracted spawning period which starts in late spring and continues through summer into early autumn. During this time many individuals engage in large spawning aggregations on the

177 lower west coast of Australia. Gillanders et al. 1999b noted that all Seriola species studied to date spawned during late spring and summer but noted that this was based on very few detailed studies. The current detailed study of the reproductive biology of S. hippos provides strong support of the view that Seriola species spawn at this time.

The length at which female S. hippos typically attain maturity (L50) (831 mm) is equivalent to a total length of 888 mm, thus exceeding the minimum legal length of retention (MML) of 600 mm TL. Therefore, although many S. hippos could get harvested prior to reproduction, this is unlikely to be the case as they are not considered a particularly good eating fish. Gillanders et al. (1999b) reported a very similar length at

50% maturity for females of 834 mm FL in the closely related S. lalandi from the waters off New South Wales.

Through providing key aspects on the age, growth and reproduction of S. hippos, this study has addressed major gaps in the understanding of the biology of this species as well as increased our knowledge of the genus Seriola. As a result, each of the two hypotheses proposed in Chapter 2 can be accepted. Seriola hippos certainly displays similar growth trajectories to other Seriola species and indeed forms the deep water aggregations west of Rottnest Island for spawning purposes. The latter finding supports the anecdotal evidence provided by Baxter (1960) and the suggestion by Gillanders et al.

1999b that S. lalandi spawns in deep offshore waters.

6.2 Movement and Migration of Seriola hippos

Seriola hippos movements along the south west coastline of Australia as displayed during the present study in relation to the annual spawning aggregations are indeed consistent with migratory behaviour. Moreover, the S. hippos spawning aggregation can best be described as transient spawning aggregations sensu Domeier and Colin (1997), thus, the hypothesis that many S. hippos individuals travel large distances to visit these

178 spawning sites annually can be accepted. Consequently, this study has documented for the first time the migratory behaviours of a large carangid on the west coast of Australia. The observed migration in S. hippos appears, to a large degree, to serve as an adaptation for utilising different habitats for growth/survival and reproduction. It is likely that the S. hippos population inhabiting the south western and southern coastal waters of Australia are well mixed as individuals tagged at Rottnest Island were recaptured at the eastern most extent of this species distribution on the south coast. Tag returns indicate, however, that spatially discrete ‘southern’ and ‘northern’ adult breeding units may exist along the west coast of Australia. This limited mixing is most likely attributable to different water temperatures throughout this species distribution, in that, some individuals, i.e. those inhabiting cooler southern waters, have a preference, or requirement, to move into warmer waters for spawning whereas other do not. This is supported by the fact that S. hippos first arrive at the Rottnest Island spawning aggregations at a time when coastal water temperatures rise quickly from winter minimums to over 18 °C. Thus, these southern individuals might only migrate as far north as the Rottnest Island region to spawn before retuning south, whereas S. hippos located north of this spawning region, e.g. Jurien Bay

(200km north), experience water temperatures above 18°C all year round (Pearce et al.

1999).

As is the case with many other marine species that inhabit the waters of the lower west coast of Australia (Lenanton et al. 1991), it is the intra and inter-annual variation in the highly influential physical oceanographic processes that is responsible for much of the spawning behaviour exhibited by S. hippos in this study. For instance, Seriola hippos shows strong spatial and temporal spawning ground fidelity as many of the fish released at the spawning aggregations were recaptured at the exact same spawning site, at approximately the same time, in subsequent seasons. Such consistent spawning behaviour at these locations strongly suggests that, in contrast to many species that aggregate to

179 spawn annually, S. hippos individuals do not synchronise spawning with a set of specific environmental conditions. Instead this observed behaviour suggests that S. hippos individuals undertake a bet-hedging reproductive strategy (Slatkin 1974) that, over a long reproductive lifespan, reduces the impact of environmental variation on reproductive success by increasing the probability that some offspring will encounter conditions favourable for survival (Goodman 1984, Helfman et al. 1997). Thus, this strategy grants greater long-term fitness in a varying environment by reducing the chance of complete reproductive failure.

This bet hedging reproductive strategy can also explain the protracted nature of the

Rottnest Island spawning aggregations. It appears that this population, consisting of individuals with different fixed behaviours, has evolved to a stable equilibrium because individuals spawning at particular times throughout season have relatively equal reproductive fitnesses over their lifespans.

This study, through a credible volunteer tagging effort, has provided the most detailed description of migration in a Seriola species to date. It is highly likely that other members of the genus also undertake migrations between habitats that enhance growth and localities which offer greater reproductive success and that these movements have not yet been detected. For example, although Gillanders et al. (2001) found no apparent patterns of movement in S. lalandi, these authors did record long distance movements comparable to those made by S. hippos. The present study supports the suggestion by Gillanders et al.

(2001) that movement patterns in the potentially wide ranging S. lalandi were obscured by a wide range of tagging locations and times. In contrast, the movement patterns of S. hippos, due to the concentrated tagging effort at spawning aggregation sites, revealed migratory behaviour.

180 6.3 Post release survival of Seriola hippos

Anglers and charter boat operators targeting S. hippos in deep waters had concerns about their long term impact on the fishery and consequently instigated a study into the post release survival of this species. It was established that the total hooking mortality of

S. hippos subjected to catch-and-release angling within the Rottnest Island sportfishery is approximately 8%. Similar to many other studies concerning catch-and-release angling, most mortality associated with this sportfishery is delayed and occurs sometime after release (Arlinghuas et al. 2007). Muoneke and Childress (1994) consider that fisheries managers should view mortality rates in a catch-and-release fishery greater than 20% with some concern. However, Coggings et al. (2007) demonstrated that life-history traits, such as longevity, growth and reproductive output, can strongly influence population responses to discard mortality (essentially equivalent to post release mortality) and recommend that such traits, together with the existing fishing mortality, should be considered when determining an acceptable catch-and-release mortality rate for a species. These authors established that discard mortality rates of less that 20% can often represent a hidden mortality source that may significantly increasing the likelihood of recruitment and growth overfishing, and decreasing fishery efficiency (through biomass loss), particularly when the fishing mortality rate is high. Thus, managers should decide on acceptable catch-and- release mortality rates on a species specific basis. The current level of mortality associated with the S. hippos catch-and-release fishery is well below 20%, in addition the schools of fish targeted by fishers are composed of many thousands of individuals and make up only a few of the known aggregations in the area. Furthermore, fishing effort at these locations is limited by the often strong prevailing winds, long distances from shore and the exertion placed on the angler by these powerful fish. Thus, given that the current fishing mortality is low, mortality associated with the catch-and-release sportfishery is likely to have little effect on the S. hippos population at present.

181 This study demonstrated that S. hippos exhibit a typical teleost neuroendocrine stress response associated with catch-and-release. Recovering levels of elevated plasma cortisol and lactate in S. hippos were consistent in duration with the general trend documented for other species for which complete recovery occurred within 24 hours post release. The results of this study also reflect similar trends to other such studies whereby mortality increases with capture depth due to barotrauma. Seriola hippos however, is much less susceptible to depth induced mortality than other commonly targeted W.A. species in which barotrauma has been observed, such as G. hebraicum.

Angler awareness of fish welfare issues and recognition of the need to develop appropriate handling protocols was widespread and common amongst participants of this fishery. Not only was this evident by the fact that this study was instigated by concerned stakeholders, but it was also reflected in the capture and handling techniques already employed within the fishery as well as the uptake of new information as it was revealed.

Key recommendation to fishers who wish to catch-and-release S. hippos at the Rottnest

Island aggregation sites are:

• Use artificial lures (jigs) with single barbless circle style hooks, bait fishermen

should use circle hooks with minimum offset.

• Target this species in waters as shallow as possible.

• Minimise fight times to reduce difficulties when releasing fish. However, slow

retrieval over the last 10 to 20 m in order to allow rapidly expanding swim bladder

gases to escape

• Reduced the time at the surface and release the fish as soon as possible.

• Use the ‘spear’ method to release fish deemed in good condition. Alternatively, if

the fish is deemed to be tired or has an inflated abdomen it should be released with

182 the aid of a large release weight, venting the swim bladder is considered invasive

and too slow.

• Larger individuals have a greater tendency to float upon release.

• Move to other fishing grounds if sharks predation regularly occurs.

For responsible recreational fishing and fish handling practices to become the norm researchers must provide easy to follow protocols that are based on good science and that the science behind these protocols must be explained in lay terms. One of the key components to the successful management of any catch-and-release fishery is sufficiently educated anglers.

The current collaborative project has revealed, and provided, a previously untapped wealth of knowledge and quality, cost effective, research assistance in Western Australia.

Through a high level of community engagement, this project has provided considerable benefit to both researchers and participating stakeholders. Cooperative research associations developed during this project have produced a large amount of research interest, increased stakeholder satisfaction from project input, improved understanding of research outcomes, and increased research uptake, all of which has led to increased stewardship and conservation of the S. hippos fishery and fisheries resources in general.

This research has proven that scientifically rigorous research collaborations directly engaging stakeholders are indeed possible and ultimately provide more comprehensive community extensions.

It is vitally important that mechanisms are set in place that ensure the uptake of these practices and their ongoing use once specific research projects have finished. The results of this study are currently incorporated into a 15 page booklet entitled “Catching and caring for Samson Fish (Seriola hippos)” published by the Fisheries Department of

Western Australia. Uptake and continual development of best handling practices is only

183 likely to occur when recreational fishers and the recreational sector, with support from researchers, takes stewardship of ongoing monitoring programs that can be used to modify protocols as new information becomes available. In the case of the S. hippos fishery, the network developed between anglers, charter boat operators, researchers and Recfishwest has been maintained and thus acts as a conduit to facilitate the ongoing development of these best handling practices as new information comes to light.

Parts of this study are applicable to any fishery. For example, with the right training recreational anglers can provide quality and cost effective data; for ongoing support and uptake of research outputs it is critical that results are conveyed to participants in a timely manner and in language they can understand; and some of the protocols can be applied to other species. Furthermore, if, as is likely, other Seriola species aggregate and sportfisheries develop around these aggregations the protocols developed in the current study are likely to provide a first step in their protection.

6.4 The unique swim bladder of Seriola hippos

It was the phenomenon of angled S. hippos to release large amounts of gas when nearing the surface that prompted initial angler concern about the survival of released S. hippos. Investigations into this release of gas revealed this physoclistic species to exhibit unique swim bladder characteristics. Seriola hippos possess a membranous tube that connects the posterior-dorsal surface of the swim bladder internally to a region under each operculum externally. This connection, termed the swim bladder vent, allows the escape of expanding swim bladder gases on rapid ascent. The presence of the swim bladder vent provides an explanation as to why the incidence of external barotrauma symptoms in S. hippos captured from the deepwater is low and why most individuals swim away strongly when released. The ability to expel excess swim bladder gases during rapid ascent whilst retaining full swim bladder function is likely to offer this semi-pelagic species

184 considerable advantages. For example, individuals are likely to be more successful when hunting physostomous prey, such as clupeids, or avoiding predators, such as sharks.

Furthermore, the capability to undertake high speed vertical movements is likely to have benefits for individuals during group spawning events, e.g. allow for the release of a greater number of gametes during broadcast spawning. In addition to S. hippos, at least two other members of the nine species of Seriola have a unique swim bladder. Future work should be directed at determining how the swim bladder vent opens, the extent of its use under natural conditions and whether it is found in other Seriola species and their relatives.

6.5 Management implications

This study suggests that S. hippos is not currently subjected to heavy fishing mortality and that both the commercial and recreational catches represent a minor proportion of the overall population. However, due to the fact that the minimum legal length of S. hippos is considerable less than the length at maturity (600 cf 888 mm TL), mortality associated fishing activities should be mindfully monitored into the future to ensure that this species is not overexploited. Furthermore, the effectiveness of future conservation measures must consider the large scale migration and spawning strategy undertaken by this species in order to ensure its protection. Overexploitation of the S. hippos spawning aggregations near Rottnest Island, in such close proximity to a relatively large population base, may have far reaching consequences. At the moment commercial fishers are not permitted to catch finfish within the area in which these aggregations occur

(Anon. 2007). Furthermore, protection is afforded to this species due to the prevalence of strong southerly winds at the aggregation sites throughout much of the spawning season which limits recreational and charter boat fishing effort. For example, one of the charter operators that concentrates on S. hippos during the aggregation period, and who aided the

185 current study, could only fish 46 days out the 138 days that constituted the 2004/05 season due to weather conditions. Recreational fishers will have even less of an impact due to smaller boats and the fact that effort is mainly concentrated on the weekends.

The S. hippos fishery is relatively unique in Western Australia in that the annual harvest of each the commercial and recreational sectors is similar. Given the increasing interest in the targeting of S. hippos for catch-and-release, managers must remain mindful of the need to persistently disseminate the handling protocols developed in the present study. If this does not occur, then mortality attributable to the recreational fishing will likely surpass that of the commercial catch in the future, particularly if, the commercial harvest remains at current levels, which is likely, as indicated by recent history.

Recently, anecdotal evidence has indicated that irresponsible behaviour by some fishers may have lead to the shark problem presently seen at the Hillarys aggregation site.

If the S. hippos sportfishery is to remain well managed then consideration should be given to introducing a requirement for charter boat operators to undertake species specific fish handling accreditation before participating in this fishery. If S. hippos is deemed worthy of management attention in the future, then managers should consider a spatial closure during the spawning season at the Hillarys Barge aggregations site. A fishing closure at this location would allow a proportion of S. hippos to undertake undisturbed spawning activities adjacent to a highly populated coastal area at the same time as reducing the apparently high catch-and-release mortality due to shark predation. A sustainable catch- and-release S. hippos fishery near the Perth metropolitan area may also offer some protection for currently over exploited local demersal species such as G. hebraicum.

Fishing effort directed at highly exploited species is likely to be reduced as some fishers will instead spend time and effort targeting S. hippos.

Whilst this study, was focussed on the S. hippos fishery around Rottnest Island the major management implications of this work are relevant to recreational fisheries

186 worldwide in that it has clearly demonstrated the value of collaborative projects. For example, this study demonstrates that robust data, producing valuable scientific insights, can be collected during such collaborations. In addition, the associations developed between researchers and stakeholders led to the ready uptake of the protocols developed during the partnership. Indeed projects of this nature would not be possible without this type of collaborative approach.

187 References

ABARE, 2006. Australian Fisheries Statistics 2005. Australian Bureau of Agricultural and Resource Economics, Canberra.

Alerstam, T., Hedenström, A. and Akesson, S., 2003. Long-distance migration: evolution and determinants. Oikos 103: 247-260.

Alexander, R. McN., 1966. Physical aspects of swim bladder function. Biological Reviews 41: 141-176.

Allen, G. R., 1997. Marine Fishes of Tropical Australian and South-east Asia: A Field Guide for Anglers and Divers. Western Australian Museum, Perth.

Andaloro, F. and Pipitone, C., 1997. Food and feeding habits of the amberjack, Seriola dumerili in the Central Mediterranean Sea during the spawning season. Cahiers de Biologie Marine 38: 91-96.

Anon., 2007. Managing the recreational catch of demersal scalefish in the west coast. Fisheries Management Paper No. 225. Department of Fisheries Western Australia., Perth.

Anon., 2008a. Recreational fishing guide, West Coast region. Department of Fisheries Western Australia., Perth.

Anon., 2008b. Recreational fishing guide, Gascoyne region. Department of Fisheries Western Australia., Perth.

Anon., 2008c. Recreational fishing guide, Pilbara/Kimberley region. Department of Fisheries Western Australia., Perth.

Anon., 2008d. Recreational fishing guide, and South Coast region. Department of Fisheries Western Australia., Perth.

Arlinghaus, R. and Cooke, S. J., 2005. Global impact of recreational fisheries. Science 307: 1561-1562.

Arlinghaus, R., Mehner, T. and Cowx, I. G., 2002. Reconciling traditional inland fisheries management and sustainability in industrialised countries, with emphasis on Europe. Fish and Fisheries 3: 261-316.

Arlinghaus, R., Cooke, S. J., Lyman, J., Policansky, D., Schwab, A., Suski, C. D., Sutton, S. G. and Thorstad, E. B., 2007. Understanding the complexity of catch-and- release in recreational fishing: an integrative synthesis of global knowledge from historical, ethical, social, and biological perspectives. Reviews in Fisheries Science 15: 75-167.

Arnold, G.P., 2001. Fish migration, horizontal. In: Steele, J., Thorpe, S. and Turekian, K. (Eds) Encyclopaedia of Ocean Sciences. Academic Press, London. pp. 947-956.

188 Arnold, G. P., and Greer-Walker, M., 1992. Vertical movements of cod (Gadus morhua L.) in the open sea and the hydrostatic function of the swimbladder. ICES Journal of Marine Science 49: 357-372.

Ayvazian, S. G., and Hyndes, G. A., 1995. Surf zone fish assemblages in south- western Australia: do adjacent nearshore habitats and the warm Leeuwin Current influence the characteristics of the fish fauna? Marine Biology 122: 527-36.

Ayvazian, S. G., Wise, B. S. and Young, G. C., 2002. Short-term hooking mortality of tailor (Pomatomus saltatrix) in Western Australia and the impact on yield per recruit. Fisheries Research 58: 241-248.

Bain, K., 2007. Sambos. In: Krantz, D. (Ed.), National Australian Fishing Annual 13: 163-175.

Baker, T. T., Lafferty, R. and Quinn, T. J., 1991. A general growth model for mark- recapture data. Fisheries Research 11: 257-281.

Ball, J.N., 1961. On the brown trout of Llyn Tegid. Proceedings of the Zoological Society of London 137: 599-622.

Ball, E.G., Strittmatter, C. F. and Cooper, O., 1955. Metabolic studies on the gas gland of the swim bladder, Biological Bulletin 108: 1-17.

Bartholomew, A. and Bohnsack, J. A., 2005. A review of catch-and-release angling mortality with implications for no-take reserves. Reviews in Fish Biology Fisheries 15: 129-154.

Barton, B.A., 2002. Stress in fishes: a diversity of responses with particular reference to changes in circulating corticosteroids. Integrative and Comparative Biology 42: 517-525.

Baxter, J.L., 1960. A study of yellowtail Seriola dorsalis (Gill). State of Californian Department of Fish and Game, Fish Bulletin 110. 91pp.

Beamish, R. J., 1979. Differences in the age of Pacific hake (Merluccius productus) using whole otoliths and sectioned otoliths. Journal of the Fisheries Research Board of Canada 36: 141-151.

Beamish, R. J. and McFarlane, G. A., 1987. Current trends in age determination methodology. In: Summerfelt, R. C. and Hall, G. E. (Eds). The age and growth of fish. Iowa State University Press, Iowa. pp. 15-42.

Beckman, D.W. and Wilson, C.A., 1995. Seasonal timing of opaque zone formation in fish otoliths. In: Secor, D.H., Dean, J.M., Campana, S.E. (Eds) Recent developments in fish otolith Research. University of South Carolina Press, Columbia. pp 27-44.

Berenbrink, M., Koldkjær, P., Kepp, O. and Cossins, A. R., 2005. Evolution of oxygen secretion in fishes and the emergence of a complex physiological system. Science 307: 1752-1757.

189 Berthold, P., 2001. Bird migration: a general survey, second edition. Oxford University Press, Oxford. 253 pp.

Bodeker, P. and Mulgrave, A., 1980. Sandgropers Guide to Better Fishing. Sixth Edition. Westsport Agencies, Subiaco.

Brewer, D. T. , Blaber, S. J. M., Milton, D. A., and Salini, J. P., 1994. Aspects of the biology of Caranx bucculentus (Teleostei: Carangidae) from the Gulf of Carpentaria, Australia. Australian Journal of Marine and Freshwater Research 45: 413-425.

Brill, R., Magel, C., Davis, M., Hannah, R. and Rankin, P., 2008. Effects of rapid decompression and exposure to bright light on visual function in black rockfish (Sebastres melanops) and pacific halibut (Hippoglossus stenolepis). Fishery Bulletin 106: 427-437.

Broadhurst, M. K., Gray, C. A., Reid, D. D., Wooden, M. E. L., Young, D. J., Haddy, J. A. and Damiano, C., 2005. Mortality of key fish species released by recreational anglers in an Australian . Journal of Experimental Marine Biology and Ecology 321: 171-179.

Brown-Peterson, N. J., Overstreet, R. M., Lotz, J. M., Franks, J. S., and Burns, K. M., 2001. Reproductive biology of cobia, Rachycentron canadum, from coastal waters of the southern United States. Fishery Bulletin 99: 15-28.

Bruesewitz, R. E., Coble, D. W., and Copes, F., 1993. Effects of deflating the expanded swim bladder on survival of burbot. North America Journal of Fisheries Management 13: 346-348.

Buckworth, R. C., 1999. Age structure of the commercial catch of Northern Territory narrow-barred Spanish mackerel. Fisheries Research and Development Corporation Report, Canberra, Australia.

Burton, M. L., Brennan, K. J., Munoz, R. C., and Parker Jr, R. O., 2005. Preliminary evidence of increased spawning aggregations of mutton snapper (Lutjanus analis) at Riley’s Hump two years after the establishment of the Tortugas South Ecological Reserve. Fisheries Bulletin 103: 404-406.

Campana, S.E. 2001. Accuracy, precision and quality control in age determination, including a review of the use and abuse of age validation methods. Journal of Fish Biology 59: 197-242.

Campbell, P. M., Pottinger T. G. and. Sumpter J. P., 1992. Stress reduces the quality of gametes produced by rainbow trout. Biology of Reproduction 47: 1140-1150.

Cappo, M., Eden, P., Newman, S. J., and Robertson, S., 2000. A new approach to validation of the periodicity and timing of opaque zone formation in the otoliths of 11 Lutjanus species from the central Great Barrier Reef. Fishery Bulletin 98: 474- 488.

190 Caputi, N., Fletcher, W. J., Pearce, A. and Chubb, C. F., 1996. Effect of the Leeuwin Current on the recruitment of fish and invertebrates along the Western Australian coast. Marine and Freshwater Research 47: 147-155.

Carter, J., and Perrine, D., 1994. A spawning aggregation of dog snapper, Lutjanus jocu (Pisces: Lutjanidae) in Belize, Central America. Bulletin of Marine Science 55: 228-234.

Castro, J. J., Santiago, J. A. and Santana-Ortega, A. T., 2002. A general theory on fish aggregation to floating objects: An alternative to the meeting point hypothesis. Reviews in Fish Biology and Fisheries 11: 255-277.

Cerrato, R.M., 1990. Interpretable statistical tests for growth comparisons using paramters in the von Bertalanffy equation. Canadian Journal of Fisheries and Aquatic Science 47: 1416-1426.

Chapman, L., Blias, A. and Gooday, P., 2001. Incorporating the recreational sector in the management of Australia’s Commonwealth fisheries. ABARE report to the Fisheries Resources Research Fund.

Chen, B. N., Qin, J. G., Carragher, J. F., Clarke, S. M., Kumar, M. S. and Hutchinson W. G., 2007. Deleterious effects of food restrictions in yellowtail kingfish Seriola lalandi during early development. Aquaculture 271: 326-335.

Claydon, J., 2004. Spawning aggregations of coral reef fishes: characteristics, hypotheses, threats and management. Oceanography and Marine Biology: An Annual Review 42: 265-302.

Clearwater, S. J. and Pankhurst, N. W., 1997. The response to capture and confinement stress of plasma cortisol, plasma sex steroids and vitellogenic oocytes in the marine teleost, red gurnard. Journal of Fish Biology 50: 429-441.

Coggins L. G., Catalano M. J., Allen M. S., Pine W. E., Walters C. J., 2007. Effects of cryptic mortality and the hidden costs of using length limits in fishery management. Fish and Fisheries 8: 196-210.

Coleman, F. C., Figueira, W. F., Ueland, J. S. and Crowder, L. B., 2004. The inpact of United States recreational fisheries on marine fish populations. Science 304: 1958- 1960.

Colin, P. L. and Clavijo, I. E., 1988. Spawning activity of fishes producing pelagic eggs on a shelf edge coral reef, southwestern Puerto Rico. Bulletin of Marine Science 43: 249-279.

Cooke, S. J. and Cowx, I. G., 2004. The role of recreational fishing in global fish crises. Bioscience 54: 857-859.

Cooke, S. J. and Suski, C. D., 2004. Are circle hooks effective tools for conserving freshwater and marine recreational catch-and-release fisheries? Aquatic Conservation: Marine and Freshwater Ecosystems 14: 299-326.

191 Cooke, S. J., and Philipp, D. P., 2004. Behavior and mortality of caught-and-released bonefish (Albula spp.) in Bahamian waters with implications for a sustainable recreational fishery. Biological Conservation 118: 599-607.

Cooke, S. J. and Suski, C. D., 2005. Do we need species-specific guidelines for catch- and-release recreational angling to conserve diverse fishery resources? Biodiversity and Conservation 14: 1195-1209.

Cooke, S. J. and Cowx, I. G., 2006. Contrasting recreational and commercial fishing: Search for common issues to promote unified conservation of fisheries resources and aquatic environments. Biological Conservation 128: 93-108.

Cooke, S. J. and Sneddon, L. U., 2006. Animal welfare perspectives on catch-and-release recreational angling. Applied Animal Behaviour Science 104: 176-198.

Cooke, S. J., Philipp, D. P., Schreer J. F., and. McKinley R. S., 2000. Locomotory impairment of nesting male largemouth bass following catch-and-release angling. North American Journal of Fisheries Management 20: 968-977.

Copeland, D. E., 1969. Fine structural study of gas secretion in the physoclistous swim bladder of Fundulus heteroclitus and Gadus callarias and in the euphysoclistous swim bladder of Opsanus tau. Z. Zellforsch 93: 305-331.

Corner, E. D. S., Denton, and Forster, G. R., 1969. On the buoyancy of some deep sea sharks. Proceeding of the Royal Society 171: 415-429.

Corsini, M., Margies, P., Kondilatos, G. and Economidis, P. S., 2006. Three new exotic fish records from the SE Aegean Greek waters. Scientia Marina 70: 319-323.

Cortés, E., 1997. A critical review of methods of studying fish feeding based on analysis of stomach contents: application to elasmobranch fishes. Canadian Journal of Fisheries and Aquatic Science 54: 726-738.

Costello, M.J., 1990. Predator feeding strategy and prey importance: a new graphical analysis. Journal of Fish Biology 36: 261-263.

Cresswell, G. R., 1980. Satellite-tracked buoys in the Eastern Indian Ocean. In: Proceedings of the Fourteenth International Symposium on Remote Sensing of the Environment, 23-30 April 1980, San Jose, Costa Rica. Environmental Research Institute of Michigan, Ann Arbor, Michigan. pp. 531-541.

Cresswell, G. R. and Peterson, J. L., 1993. The Leeuwin Current south of Western Australia. Australian Journal of Marine and Freshwater Research 44: 285-303.

Cresswell, G.R. and Griffin, D.A., 2004. The Leeuwin Current, eddies and sub-Antarctic waters off south-western Australia. Marine and Freshwater Research 55: 267-276.

Cresswell, G. R., Boland, F. M., Peterson, J. L. and Wells, G. S., 1989. Continental shelf currents near the Abrolhos Islands, Western Australia. Australian Journal of Marine and Freshwater Research 40: 113-128.

Cusack, R. and Roennfeldt, M., 1988. Fishing the Wild West. St George Books, Perth.

192 de Lestang, P., Griffin, R. K. and Allsop, Q. A., 2004. Assessment of the post-release survival and stress physiology of barramundi (Lates calcarifer). Fishery Report No. 73. Northern Territory Government. 45 pp.

Diggles, B. K. and Ernst, I., 1997. Hooking mortality of two species of shallow water reef fish caught by recreational angling methods. Marine and Freshwater Research 48: 479-483.

Dingle, H., 1996. Migration: the biology of life on the move. Oxford University Press, New York. 474pp.

Dingle, H. and Drake, V. A., 2007. What is migration? BioScience 57: 113-121.

Dingle, H. and Holyoak, M., 2001. The evolutionary ecology of movement. In: Fox, C. W., Roff, D. A. and Fairbairn, D.J. (Eds.) Evolutionary Ecology: Concepts and Case Studies. Oxford University Press, New York. pp 247-260.

Domeier, M. L. and Colin, P. L., 1997. Tropical reef fish spawning aggregations: defined and reviewed. Bulletin of Marine Science 60: 698-726.

Ewing, G., Welsford, D., Jordan, A. R., and Buxton, C.D., 2003. Validation of age and growth estimates using thin otolith sections from the purple wrasse, Notolabrus fucicola. Marine and Freshwater Research, 54: 985-993.

Fairclough, D. V., Dimmlich, W. F., and Potter, I. C., 2000. Length and age compositions and growth rates of the Australian herring Arripis georgiana in different regions. Marine and Freshwater Research 51: 631-640.

FAO, 1998. FAO yearbook fisheries statistics. Volume 82. Food and Agriculture Organisation of the United Nations, Rome.

FAO, 2007. The state of world fisheries and aquaculture 2006, Food and Agriculture organisation of the United Nations, Rome.162 pp.

Farley, J. H. and Davis, T. L. O., 1998. Reproductive dynamics of southern bluefin tuna, Thunnus maccoyii. Fisheries Research 96: 223-236.

Farmer, B. M., French, D. J. W., Potter, I. C., Hesp, S. A., and Hall, N. G., 2005. Determination of the biological parameters for managing the fisheries for Mulloway and Silver Trevally in Western Australia. Fisheries Research and Development Corporation Report, FRDC Project No. 2002/180, Canberra, Australia.

Farrell A. P., Gallaugher, P. E. and Routledge, R., 2001. Rapid recovery of exhausted adult coho salmon after commercial capture by troll fishing. Canadian Journal of Fisheries and Aquatic Science 58: 2319-2324.

Feathers, M. G. and Knable, A. E., 1983. Effects of depressurization upon largemouth bass. North American Journal of Fisheries Management 3: 86-90.

193 Feng, M., Meyers, G., Pearce, A. and Wijffels, S., 2003. Annual and interannual variations of the Leeuwin Current at 32°S. Journal of Geophysical Research, 108(11), 3355, doi:10.1029/2002JC001763.

Ferguson, R.A., and Tufts, B.L., 1992. Physiological effects of brief air exposure in exhaustively exercised rainbow trout (Oncorhynchus mykiss): implications for 'catch and release' fisheries. Canadian Journal of Fisheries and Aquatic Science 49: 1157-1162.

Findlay, J., 2007. Small Pelagics Fishery. In: Larcombe, J. and McLoughlin, K. (Eds.) Fishery Status Reports 2006: Status of Fish Stocks Managed by the Australian Government. Bureau of Rural Sciences, Canberra. pp. 217-220.

Francis, R. I. C. C., 1988a. Are growth parameters estimated from tagging and age-length data comparable? Canadian Journal of Fisheries and Aquatic Science 45: 936-942.

Francis, R. I. C. C., 1988b. Maximum likelihood estimation of growth and growth variability from tagging data. New Zealand Journal of Marine and Freshwater Research 22: 42-51.

Francis, R. I. C. C., 1995. An alternative mark-recapture analogue of Schnute's growth model. Fisheries Research 23: 95-111.

Franks, J. S., Warren, J. R., and Buchanan, M. V., 1999. Age and growth of cobia, Rachycentron canadum, from the northeastern Gulf of Mexico. Fishery Bulletin 97: 459-471.

Frisch, A. and Anderson, T., 2005. Physiological stress responses of two species of coral trout (Plectropomus leopardus and Plectropomus maculatus). Comparative Biochemistry and Physiology, Part A 140: 317-327.

Gaughan D. J., White K.V. and Fletcher, W. J., 2001. The links between functionally distinct adult assemblages of Sardinops sagax: larval advection across management boundaries. ICES Journal of Marine Science 58: 597-606.

Gersbach, G. H., Pattiaratchi, C. and Ivey, G. N., 1999. Upwelling on the south-west coast of Australia- source of the Capes Current? Continental Shelf Research 19: 363-400.

Gill, H. S., Wise, B. S., Potter, I. C. and Chaplin, J. A., 1996. Biannual spawning periods and resultant divergent patterns of growth in the estuarine goby Pseudogobius olorum: temperature-induced? Marine Biology 125: 453-466.

Gillanders, B. M., Ferrell, D. J., Andrew, N. L., 1999a. Ageing methods for yellowtail kingfish (Seriola lalandi, Carangidae) and results from age- and size-based growth models. Fishery Bulletin 97: 812-827.

Gillanders, B. M., Ferrell, D. J., Andrew, N. L., 1999b. Size at maturity and seasonal changes in gonad activity of yellowtail kingfish (Seriola lalandi; Carangidae) in New South Wales, Australia. New Zealand Journal of Marine and Freshwater Research 33: 457-468.

194 Gillanders, B.M., Ferrell, D.J., and Andrew, N.L., 2001. Estimates of movement and life- history parameters of yellowtail kingfish (Seriola lalandi): how useful are data from a cooperative tagging programme? Marine and Freshwater Research 52:179- 192.

Gomon, M.F., Glover, J. C. M., and Kuiter, R. H., (Eds), 1994. The Fishes of Australia’s South Coast. Flora and Fauna of South Australia Handbooks Committee. State Printer, Adelaide.

Goodman D., 1984. Risk spreading as an adaptive strategy in iteroparous life histories. Theoretical Population Biology 25: 1-20.

Graham, R.T. and Castellanos, D. W., 2005. Courtship and spawning behaviours of carangid species in Belize. Fishery Bulletin 103: 426-432.

Grant, E. M., 1991. Grant’s Fishes of Australia. E. M. Grant Pty Limited, Redcliffe, Queensland.

Grant, E. M., 1999. Grant’s Guide to Fishes. E. M. Grant Pty Limited, Scarborough, Queensland.

Green, J., 2005. The application of remote sensing to locate shipwrecks. In: Jones, I. S. F. and Zhang, S. (Eds), Australian-Chinese Ocean Science & Technology - The Proceedings, http://hdl.handle.net/2123/1420. pp 23-27.

Gregory, W. K., 1933. Fish skulls: a study of the evolution of natural mechanisms. Transactions of the American Philosophical Society 23(2): 75-481.

Gunn, J. S., 1990. A revision of selected genera of the family Carangidae (Pisces) from Australian waters. Record of the Australian Museum, Supplement 12: 1-77.

Hall, N. G., Hesp, S. A., and Potter, I. C., 2004. A Bayesian approach for overcoming inconsistencies in mortality estimates using, as an example, data for Acanthopagrus latus. Canadian Journal of Fisheries and Aquatic Science 61: 1202-1211.

Harden-Jones, F. R. and Scholes, P., 1985. Gas secretion and resorption in the. swimbladder of cod Gadus morhua. Journal of Comparative Physiology 155: 319- 331.

Harris, P. J., 2004. Age, growth and reproduction of greater amberjack, Seriola dumerili, in the southwestern North Atlantic. Analytical Report of Marine Resource Monitoring, Assessment and Prediction Program. South Carolina Department of Natural Resources, Charleston.

Heibo, E., Magnhagen, C., and Vollestad, L.A., 2005. Latitudinal variation in life-history traits in Eurasian perch. Ecology, 86: 3377-3386.

Helfman, G., Collette, B. and Facey, D., 1997. The diversity of fishes. Blackwell Science, Malden. 528 pp.

195 Henry, G. W. and Lyle, J. M. (Eds), 2003. The National Recreational and Indigenous Fishing Survey. FRDC Project No. 99/158. Australian Government Department of Agriculture, Fisheries and Forestry, Canberra.

Herskin, J. and Steffensen, J. F., 1998. Energy savings in sea bass swimming in a school: measurements of tail beat frequency and oxygen consumption at different swimming speeds. Journal of Fish Biology 53: 366-376.

Hesp, S. A., Potter, I. C., and Hall, N. G., 2002. Age and size compositions, growth rates, reproductive biology and habitats of the West Australian dhufish (Glaucosoma hebraicum) and their relevance to the management of this species. Fishery Bulletin 100: 214-227.

Heyman, H. D., Kjerfve, B., Graham, R. T., Rhodes, K. L. and Garbutt, L., 2005. Spawning aggregations of Lutjanus cyanopterus (Cuvier) in the Belize Barrier Reef over a 6 year period. Journal of Fish Biology 67: 83-101.

Heyman, H. D., Graham, R. T. Kjerfve, B. and Johannes, R. E., 2001. Whale sharks, Rhincodon typus, aggregate to feed on fish spawn in Belize. Marine Ecological Progress Series 215: 275-282.

Hill, B. J., 1990. Keynote address: Minimum legal sizes and their use in management of Australian fisheries. In: Hancock, D. A. (Ed), Australian Society for Fish Biology Workshop Proceedings. Australian Government Publishing Service, Canberra. pp. 9-50.

Hoenig, J. M., 1982. A compilation of mortality and longevity estimates for fish, molluscs, and cetaceans, with a bibliography of comparative life history studies. GSO Technical Report Reference No. 82-2.

Hoenig, J. M., 1983. Empirical use of longevity data to estimate mortality rates. Fishery Bulletin 81: 898-903.

Honebrink, R. R., 2000. A review of the biology of the family Carangidae, with emphasis on species found in Hawaiian waters. DAR Technical Report 20-01. Division of Aquatic Resources, Department of Land and Natural Resources Honolulu, (HI). 43 pp.

Holdsworth, J. and Saul, P., 2003. New Zealand billfish and gamefish tagging 2001–02. New Zealand Fisheries Assessment Report 2003/15. Ministry for Fisheries, Wellington. 39 pp.

Hollis, E. H., 1948. The homing tendency of Shad. Science 108: 332-333.

Houde, E. D., 2008. Emerging from Hjort’s shadow. Journal of Northwest Atlantic Fishery Science 41: 53-70.

Hutchings, L., Beckley, L. E., Griffiths, M. H., Roberts, M. J., Sundby, S. and van der Lingen, C., 2002. Spawning on the edge: spawning grounds and nursery areas around the southern African coastline. Marine and Freshwater Research. 53: 307- 318.

196 Hutchins, B. and Swainston, R., 2002. Sea Fishes of Southern Australia. Swainston Publishing, Perth.

Hutchins, B. and Thompson, M., 1995. The Marine and Estuarine Fishes of South-western Australia: A Field Guide for Anglers and Divers. Western Australian Museum, Perth.

Hunter E., Metcalfe J. D. and Reynolds J. D., 2003. Migration route and spawning area fidelity by North Sea plaice. Proceedings of the Royal Society Series B- Biological Sciences 270: 2097-2103.

Hunter, J. R., Lo, N. C. H. and Leong, R. J. H., 1985. Batch fecundity in multiple spawning fishes. In:Lasker, R. (Ed), An egg production method for estimating spawning biomass of pelagic fish: application to the northern anchovy, Engraulis mordax (.), p. 67-77. Dep. Commer., NOAA Tech. Rep. NMFS 36.

Hutson, K. S., Smith, B. P., Godfrey, R.T., Whittington, I. D., Chambers, C. B., Ernst, I. and Gillanders, B..M., 2007a. A tagging study on yellowtail kingfish (Seriola lalandi) and Samson fish (S. hippos) in South Australian waters. Transactions of the Royal Society of South Australia 131: 128-134.

Hutson K.S., Ernst I. and Whittington I.D., 2007b. Risk assessment for parasites of Seriola lalandi (Carangidae) in South Australian sea-cage aquaculture. Aquaculture 271: 85-99.

Hynes, H. B. N., 1950. The food of sticklebacks with a review of the methods used in studies of the food of fishes. Journal of Animal Ecology 19: 36-58.

Hyslop, E.J., 1980. Stomach contents analysis – a review of methods and their application. Journal of Fish Biology 17: 411-429.

James, G. D., 1984. Trevally, Caranx georgianus Curvier: Age determination, population biology and fishery. New Zealand Ministry of Agriculture and Fisheries, Fisheries Research Bulletin No. 25.

Jerez, S., Samper, M., Santamaria, F. J., Villamandos, J., Cejas, J. and Felipe, B., 2006. Natural spawning of greater amberjack (Seriola dumerili) kept in captivity in the Canary Islands. Aquaculture 252: 199-207.

Jover, M., García, A., Tomás, A., de la Gándara, F. and Pérez. L., 1999. Growth of Mediterranean yellowtail (Seriola dumerili) fed extruded diets containing different levels of protein and lipid. Aquaculture 179: 25-33.

Kailola, P. J., Williams, M. J., Steward, P. C., Reichelt, R. E., McNee, A. and Grieve, C., 1993. Australian Fisheries Resources. Bureau of Resource Science and Fisheries Development Cooperation, Canberra.

Kearney, R. E., 1989. Keynote address: tagging - solution or problem? In: Hancock, D. A. (Ed), Proceedings No. 5 Australian Society for Fish Biology tagging workshop, Bureau of Rural Resources, Department of Primary Industries and Energy. pp. 8- 20.

197 Kennedy, J. S., 1985. Migration, behavioural and ecological. In: Rankin, M. A. (Ed.), Migration: mechanisms and adaptive significance. Contributions in Marine Science 27(supplement): 5-26.

Kerr, S. J., 2001. A review of “fizzing’ – a technique for swim bladder deflation. Ontario Ministry of Natural Resources, Fisheries Section, Peterborough, Ontario. pp 13.

Kimura, D. K., 1980. Likelihood methods for von Bertalanffy growth curve. Fisheries Bulletin 77: 765-776.

Kimura, S., Kasai, A. and Sugimoto, T., 1994. Migration of yellowtail in relation to intrusions of warm water from the Kuroshio. Fisheries Science 60: 635-641

Kishore, R. and Solomon, F., 2005. Age and Growth Studies of Caranx hippos (crevalle jack) from Trinidad Using Hard-parts. Proceedings of the Gulf and Caribbean Fisheries Institute 56: 227-239. (abstract only).

Kozul, V., Skaramuca, B., Kraljevic, M., Dulcic, J., and Glamuzina, B., 2001. Age, growth and mortality of the Mediterranean amberjack Seriola dumerili (Risso 1810) from the south-eastern Adriatic Sea. Journal of Applied Ichthyology 17: 134-141.

Labelle, M., Hampton, J., Bailey, K., Murray, T., Fournier, D. A. and Sibert, J. R., 1993. Determination of age and growth of South Pacific albacore (Thunnus alalunga) using three methodologies. Fishery Bulletin 91: 649-663.

Laroche, W. A., Smith-Vaniz W. F. and Richardson S.L., 1984. Carangidae: Development. In: Moser, H. G., Richards, W. J., Cohen, D. M., Fahay M. P., Kendall A. W., and Richardson S.L. (Eds.), Ontogeny and systematics of fishes. Special Publication 1, American Society of Ichthyologists and Herpetologists: 510- 522.

Leggett, W. C., 1977. The ecology of fish migrations. Annual Review of Ecology and Systematics 8: 285-308.

Leis, J. M., 1991. The pelagic stage of reef fishes. In: Sale, P. F. (Ed), The ecology of fishes on coral reefs. Academic Press, San Diego.

Lenanton, R. C., Joll, L., Penn, J. and Jones, K., 1991. The influence of the Leeuwin Current on coastal fisheries of Western Australia. In: Pearce, A. F. and Walker, D. I. (Eds), The Leeuwin Current: An Influence on the Coastal Climate and Marine Life of Western Australia. Journal of the Royal Society of Western Australia 74: 101-114.

Lewis, P. D. and Mackie, M. C., 2002. Methods used in the collection, preparation and interpretation of narrow-barred Spanish mackerel (Scomberomorus commerson) otoliths for a study of age and growth in Western Australia. Fisheries Research Report No.143. Department of Fisheries, Perth.

Lewin, W. C., Arlinghaus, R. and Mehner, T., 2006. Documented and potential biological impacts of recreational angling: Insights for conservation and management. Reviews in Fisheries Science 14: 305-367.

198

Lin, P. L. and Shao, K. T., 1999. A review of the Carangid fishes (Family Carangidae) from Taiwan with descriptions of four new records. Zoological Studies 38: 33-68.

Lord, D. A. and Hillman, K., 1995. Perth coastal waters study: summary report. Report to the Water Authority of Western Australia. 134 pp.

Love, G. and Langenkamp, D., 2003. Australian aquaculture: industry profiles for selected species. ABARE eReport no 03.8 prepared for the Fisheries Resources Research Fund. Canberra: Australian Bureau of Agriculture and Resource Economics. 128 pp.

Love, M. S., Mecklenburg, C. W., Mecklenburg, T. A. and Thorsteinson, L. K., 2005. Resource inventory of marine and estuarine fishes of the West Coast and Alaska: a checklist of North Pacific and Arctic Ocean species from Baja California to the Alaska-Yukon border. OCS Study MMS 2005-030 and USGS/NBII 2005-001, US Dept Interior, US Geological Survey, Biological Resources Division, Seattle, WA. pp 1-276.

Lowerre-Barbieri, S. K., Vose F. E, and. Whittington J. A., 2003. Catch-and-release fishing on a spawning aggregation of common snook: Does it affect reproductive output? Transaction of the North American Fisheries Society 132: 940-952.

Lucy, J. and Davy, K., 2000. Benefits of angler-assisted tag and release programs. Fisheries 25(4): 18-23.

Mackie, M. C. and Lewis, P. D., 2001. Assessment of gonad staging systems and other methods used in the study of the reproductive biology of narrow-barred Spanish mackerel, Scomberomorus commerson, in Western Australia. Fisheries Research Report No.136. Department of Fisheries, Perth.

Manooch, C. S. and Potts, J. C., 1997a. Age, growth and mortality estimates of greater amberjack, Seriola dumerili, from the U.S. Gulf of Mexico headboat fishery. Bulletin of Marine Science 61: 671-683

Manooch, C. S. and Potts, J. C., 1997b. Age, growth and mortality of greater amberjack from the southeastern United States. Fisheries Research 30: 229-240.

Mant, J. C., Moran, M. J., Newman, S. J., Hesp, A. S., Hall, N. G., and Potter, I. C., 2006. Biological characteristics and mortality of Pentapodus vitta, an abundant by-catch species of prawn trawling and recreational fishing in a large subtropical embayment. Fishery Bulletin 104: 512-520.

Marino, G., Mandich, A., , A., Andaloro, F., Porrello, S., Finoia, M. G. and Cevasco, F., 1995. Aspects of reproductive biology of the Mediterranean amberjack (Seriola dumerili Risso) during the spawning period. Journal of Applied Ichthyology 11: 9-24.

Marshall, S. and Elliott, M., 1997. A comparison of univariate and multivariate numerical and graphical techniques for determining inter- and intraspecific feeding relationships in estuarine fish. Journal of Biology 51: 526-545.

199 Matallanas, J., Casadevall, M., Carrassón, M., Boix, J. and Fernández, V., 1995. The food of Seriola dumerili (Pisces: Carangidae) in the Catalan Sea (western Mediterranean). Journal of the Marine Biological Association of the United Kingdom 75: 257-260.

Matthews, L. H., 1978. The Natural History of the Whale. Weidenfeld and Nicholson, London. pp 219.

Maxwell, J. G. H. and Cresswell, G. R., 1981. Dispersal of tropical marine fauna to the Great Australian Bight by the Leeuwin Current. Marine and Freshwater Research 32: 493-500.

Maynard Smith, J., 1982. Evolution and the Theory of Games. Cambridge University Press, Cambridge. 224 pp.

Mazzola, A., Favaloro, E. and Sara, G., 2000. Cultivation of the Mediterranean amberjack, Seriola dumerili (Risso, 1810), in submerged cages in the Western Mediterranean Sea. Aquaculture 181: 257-268.

McCune, A. R. and Carlson, R. L., 2004. Twenty ways to lose your bladder: common natural mutants in zebrafish and widespread convergence of swim bladders among teleost fishes. Development and Evolution 6: 246-259.

McLeay, L. J., Jones, G. K. and Ward, T. M., 2002. National Strategy for the Survival of Released Line-Caught Fish: a Review of Research and Fishery Information (FRDC Project 2001/101). South Australian Research and Development Institution: Henley .

Meka, J. M., 2004. The influence of hook type, angler experience, and fish size on injury rates and the duration of capture in an Alaskan catch-and-release rainbow trout fishery. North American Journal of Fisheries Management 24: 1309-1321.

Metcalfe, J. D., Hunter, E. and Buckley, A. A., 2006. The migratory behaviour of North Sea plaice: Currents, clocks and clues. Marine and Freshwater Behaviour and Physiology 39: 25-36.

Michale, V., Maricchiolo, G. and Genovese, L., 1999. The reproductive biology of the amberjack, Seriola dumerilii (Risso 1810). I. Oocyte development in captivity. Aquaculture Research 30: 349-355.

Milligan, C. L. and Girard, S. S., 1993. Lactate metabolism in trout. Journal of Experimental Biology 180:175-193.

Milligan, C. L., Hooke, G. B., and Johnson, C., 2000. Sustained swimming at low velocity following a bout of exhaustive exercise enhances metabolic recovery in rainbow trout. The Journal of Experimental Biology 203: 921-926.

Mommsen, T. P., Vijayan, M. M. and Moon, T. W., 1999. Cortisol in teleosts: dynamics, mechanisms of action, and metabolic regulation. Reviews in Fish Biology and Fisheries 9: 211-268.

200 Moran, M., Burton, C. and Jenke, J., 2003. Long-term movement patterns of continental shelf and inner gulf pink snapper (Pagrus auratus, Sparidae) from tagging in the Shark Bay region of Western Australia. Marine and Freshwater Research 54: 913- 922.

Moran D., Smith C. K., Gara B., Poortenaar C. W., 2007. Reproductive behaviour and early development in yellowtail kingfish (Seriola lalandi Valenciennes 1833). Aquaculture 262: 95-104.

Morgan, M. J., Wilson, C.E. and Crim L.W., 1999. Effects of stress on reproduction in Atlantic cod. Journal of Fish Biology 54: 477-488.

Moser, M. L., Auster, P. J., and Bichy, J. B., 1998. Effects of mat morphology on large Sargassum-associated fishes: observations from remotely operated vehicle (ROV) and free-floating video camcorders. Environmental Biology of Fishes 51: 391-398.

Moyle, P. B. and Cech, J. J., 2004. Fishes: An Introduction to Ichthyology (5th ed.), Prentice Hall, New Jersey.

Muhling, B. A., and Beckley, L. E., 2007. Seasonal variation in horizontal and vertical structure of larval fish assemblages off south-western Australia, with implications for larval transport. Journal of Plankton Research 29: 967-983.

Muhling, B. A., Beckley, L. E., Koslow, J. A., Pearce, A. F., 2008a. Larval fish assemblages and water mass structure off the oligotrophic south-western Australian coast. Fisheries Oceanography 17: 16-31.

Muhling, B. A., Beckley, L. E., Gaughan, D. J., Jones, C. M., Miskiewicz, A. G. and Hesp, S. A., 2008b. Spawning, larval abundance and growth rate of Sardinops sagax off southwestern Australia: influence of an anomalous eastern boundary current. Marine Ecology Progress Series 364: 157-167.

Muoneke, M. I., and. Childress W. M. 1994. Hooking mortality: a review for recreational fisheries. Reviews in Fisheries Science 2:123-156.

Mylonas, C.C., Papandroulakis, N., Smboukis, A. and Divanach, P., 2004. Induction of spawning of cultured greater amberjack (Seriola dumerili) using GnRHa implants. Aquaculture 237: 141-154.

Nakada, M., 2002. Yellowtail culture development and solutions for the future. Reviews in Fisheries Science 10: 559-575.

Nelson, J. S., 2006. Fishes of the world. John Wiley and Sons, Inc. New York. 4th edition.

Nelson, J. S., Crossman, E. J., Espinosa-Pérez, H., Findley, L. T., Gilbert C. R., Lea R. N. and Williams J. D., 2004. Common and scientific names of fishes from the United States, Canada, and Mexico. Sixth Edition. Special Publication 29. American Fisheries Society, Bethesda, Maryland. 386 pp.

Nemeth, R. S., Blondeau, J., Herzlieb, S., and Kadison, E., 2007. Spatial and temporal patterns of movement and migration at spawning aggregations of red hind,

201 Epinephelus guttatus, in the U.S. Virgin Islands. Environmental Biology Fishes 78: 365-381.

Okata, A., 1976. Ecological studies on the biological production of young amberfish community in the Sendai Bay-II, Relationships between food chains and fish fauna. Bulletin of the Japanese Society of Scientific Fisheries 42: 29-44.

Olla, B. L., Davis, M. W. and Schreck, C. B., 1998. Temperature magnified postcapture mortality in adult sablefish after simulated trawling. Journal of Fish Biology 53: 743-751.

Ortiz, M., Prince, E. D., Serafy, J. D., Holts, D. B., Davy, K. B., Pepperell, J. G., Lowry, M. B., Holdsworth, J. C., 2003. A global overview of the major constituent-based billfish tagging programs and their results since 1954. Marine and Freshwater Research 54: 489-507.

Pankhurst, N. W. and Sharples, D. F., 1992. Effects of capture and confinement on plasma cortisol concentrations in the snapper, Pagrus auratus. Marine and Freshwater Ressearch 43: 345-356.

Pankhurst, N. W. and Dedual, M., 1994. Effects of capture and recovery on plasma levels of cortisol, lactate and gonadal steroids in a natural population of rainbow trout. Journal of Fish Biology 45: 1013-1025.

Parker, G. A., 1992. The evolution of sexual dimorphism in fish. Journal of Fish Biology 41 (Supplement B): 1-20.

Parker, S. J., Olsom, J. M., Rankin, P. S. and Malvitch, J. S., 2008. Patterns in vertical movements of black rockfish Sebastes melanops. Aquatic Biology 2: 57-65.

Parmenter, C. J., 1983. Reproductive migration in the Hawksbill Turtle (Eretmochelys imbricata). Copeia 1983: 271-273.

Parrish, F. A., and Moffitt, R. B., 1993. Subsurface fish handling to limit decompression effects on deepwater species. Marine Fisheries Review 54: 29-32.

Pauly, D., 1980. On the interrelationships between M, growth parameters, and mean environmental temperature in 175 fish stocks. Journal du Conseil International pour l’Exploration de la Mer 39: 175-192.

Paxton, J. R., Hoese, D. F., Allen, G. R. and Hanley, J. E., 1989. Zoological catalogue of Australia. Volume 7, Pisces. Petromyzontidae to Carangidae. Australian Government Publishing Service, Canberra.

Pearce, A. F., and Phillips, B. F., 1988. ENSO events, the Leeuwin Current, and recruitment to the western rock lobster fishery. Journal du Conseil International pour l'Exploration de la Mer 45 :13-21.

Pearce, A. F. and Griffiths, R. W., 1991. The mesoscale structure of the Leeuwin Current: a comparison of laboratory models and satellite imagery. Journal of Geophysical Research 96: 16739-16757.

202 Pearce, A. F. and Pattiaratchi, C.,1999. The Capes Current: a summer countercurrent flowing past Cape Leeuwin and Cape Naturaliste, Western Australia. Continental Shelf Research 19: 401-420.

Pearce, A. F., Rossbach, M., Tait, M. and Brown, R., 1999. Sea temperature variability of Western Australia 1990 to 1994. Fisheries Research Report No. 111. Department of Fisheries, Perth. 45pp.

Pearce, A. F., Lynch, M. J. and Hanson, C. E., 2006. The Hillarys transect (1): seasonal and cross-shelf variability of physical and chemical water properties off Perth, Western Australia, 1996–98. Continental Shelf Research 26: 1689-1729.

Pelster B., 1998. Buoyancy. In: Evans, D. H. (Ed), The Physiology of Fishes. Second Edition. CRC Press, Boca Raton. pp 25-42.

Peter, B. N., Sreeraji, P. and Vimal Kumar, K.G., 2005. Structure and variability of the Leeuwin Current in the south eastern Indian Ocean. Journal of Indian Geophysical Union 9: 107-119.

Phleger, C.F. 1998. Buoyancy in marine fishes: direct and indirect role of lipids American Zoologist 38: 321-330.

Philip, R. B., 1974. A review of blood changes associated with compression- decompression: relationship to decompression sickness. Undersea Biomedical Research 1: 117-150.

Philipp, D. P., Toline, C. A.,. Kubacki, M. F., Philipp, D. B. F. and. Phelan F. J. S., 1997. The impact of catch-and-release angling on the reproductive success of smallmouth bass and largemouth bass. North American Journal of Fisheries Management 17: 557-567.

Pilliod, D. S., Peterson, C. R., Ritson, P. I., 2002. Seasonal migration of Columbia spotted frogs (Rana luteiventris) among complementary resources in a high mountain basin. Canadian Journal of Zoology 80: 1849-1862.

Pizzicori, P., Castriota, L., Marino, G. and Andaloro, F., 2000. Seriola carpenteri: a new immigrant in the Mediterranean Sea from the Atlantic Ocean. Journal of Fish Biology 57: 1335-1338.

Poortenaar, C.W., Hooker, S.H., and Sharp, N.J., 2001. Assessment of yellowtail kingfish (Seriola lalandi lalandi) reproductive physiology, as a basis for aquaculture development. Aquaculture 201: 271-286.

Poortenaar, C., Jeffs, A., Heath., P. and Hooker, S., 2003. Commercial opportunities for kingfish aquaculture in Northland. NIWA Client Report: AKL2003-026. National Institute of Water and Atmospheric Research Ltd, Auckland.

Pottinger, T. G., 1998. Changes in blood cortisol, glucose and lactate in carp retained in anglers’ keepnets. Journal of Fish Biology 53: 728-742.

Pough, F. H., Heiser, B. H. and McFarland, W. N., 1996. Vertebrate life. Prentice-Hall International Limited, London.

203

Prokop, F., 2006. Australian Fishing Guide. Australian Fishing Network, Croydon South, Victoria.

Reed, D. L., Carpenter, K. E. and deGravelle, M. J., 2002. Molecular systematics of the Jacks (Perciformes: Carangidae) based on mitochondrial cytochrome b sequences using parsimony, likelihood, and Bayesian approaches. Molecular Phylogenetics and Evolution 23: 513-524.

Rhodes, K. L. & Sadovy, Y., 2002. Temporal and spatial trends in spawning aggregations of camouflage grouper, Epinephelus polyphekadion, in Pohnpei, Micronesia. Environmental Biology of Fishes 63: 27-39.

Richards, W. J., and Dove, G. R., 1971. Internal development of young tunas in the genera Katsuwonus, Euthynnus, Auxis, and Thunnus (Pisces, Scombridae). Copeia 1971: 72-78.

Ricker, W. E., (1975). Computation and interpretation of biological statistics of fish populations. Bulletin of the Fisheries Research Board of Canada 191.

Robichaud, D. and Rose G. A., 2001. Multiyear homing of Atlantic cod to a spawning ground. Canadian Journal of Fisheries and Aquatic Sciences 58: 2325-2329.

Rombough, P. J., 1997. The effects of temperature on embryonic and larval development. In: Wood, C.M. and McDonald, D.G. (Eds.), Global Warming. Implications for Freshwater and Marine Fish. Cambridge University Press, Cambridge. pp. 177- 223.

Rooker, J. R., Turner, J. P. and Holt, S. A., 2006. Trophic ecology of Sargassum- associated fishers in the Gulf of Mexico determined from stable isotopes and Fatty acids. Marine Ecology Progress Series 313: 249-259.

Root, R. W., 1931. The respiratory function of the blood of marine fishes. Biological Bulletin 61: 427-456.

Ruello, N. V., 1975. Geographical distribution, growth and breeding migration of the eastern Australian king prawn Penaeus plebejus Hess. Australian Journal of Marine and Freshwater Research 26: 343-354.

Rummer, J. L. and Bennett, W. A., 2005. Physiological effects of swim bladder overexpansion and catastrophic decompression on red snapper, Lutjanus campechanus. Transactions of the American Fisheries Society 134: 1457-1470.

Sakakura, Y. and Tsukamoto, K., 1997. Age composition in the schools of juvenile yellowtail Seriola quinqueradiata associated with drifting seaweeds in the East China Sea. Fisheries Science 63:37-41.

Sala, E., Aburto-Oropereza, O., Paredes, G. and Thompson, G., 2003. Spawning aggregations and reproductive behavior of reef fishes in the Gulf of California. Bulletin of Marine Science 72: 103-121.

204 Sanderson, L. S., Chesnutt, C. R. and Lobel, P. S., 1996. Evidence for ram suspension feeding by the piscivore, Seriola dumerili (Carangidae). Environmental Biology of Fishes 46: 365-373.

Saul, P. and Holdsworth, J., 1992. Cooperative gamefish tagging in New Zealand waters, 1975–90. New Zealand Fisheries Technical Report No. 33. 24 pp.

Schaefer K. M., 1996. Spawning time, frequency, and batch fecundity of yellowfin tuna, Thunnus albacares, near Clipperton Atoll in the eastern Pacific Ocean. Fishery Bulletin 94: 66-76.

Scheer, B. T., 1939. Homing instinct in salmon. Quarterly Review of Biology 14: 408- 430.

Schill, D. J., 1996. Hooking mortality of bait-caught rainbow trout in an Idaho trout stream and a hatchery: implications for special-regulation management. North American Journal of Fisheries Management 16: 348-356.

Schmitt, R.J. and Strand, S. W., 1982. Cooperative foraging by yellowtail, Seriola lalandei (Carangidae), on two species of fish prey. Copeia 1982(3): 714-717.

Schnute, J., 1981. A versatile growth model with statistically stable parameters. Canadian Journal of Fisheries and Aquatic Science 38: 1128-1140.

Schwarz, C. J. and Seber, G. A. F., 1999. Estimating animal abundance: review III. Statistical Science 14: 427-456.

Shapiro, D. Y., Sadovy, Y. and McGhee, M. A., 1993. Size, composition, and spatial structure of the annual spawning aggregation of the red hind, Epinephelus guttatus (Pisces: Serranidae). Copeia 1993: 399-406.

Shasteen, S. P. and Sheehan, R. J., 1997. Laboratory evaluation of artificial swim bladder deflation in large-mouth bass: Potential benefits for catch-and-release fisheries. North American Journal of Fisheries Management 17: 32-37.

Skomal, G. B., 2007. Evaluating the physiological and physical consequences of capture on post-release survivorship in large pelagic fishes. Fisheries Management and Ecology 14: 81-89.

Slatkin, M., 1974. Hedging one’s evolutionary bets. Nature 250: 704-705.

Smith-Vaniz, W.F., 1984. Carangidae: Relationships. In Moser, H. G., Richards, W. J., Cohen, D. M., Fahay M. P., Kendall A. W., and Richardson S.L. (Eds.), Ontogeny and systematics of fishes. Special Publication 1, American Society of Ichthyologists and Herpetologists: 522-530.

Smith-Vaniz, W. F., 1999. Carangidae. In: Carpenter, K. E., and Niem, V. H. (Eds). The living marine resources of the Western Central Pacific. Volume 4. Bony fishes part 2 (Mugilidae to Carangidae). FAO species identification guide for fishery purposes. FAO, Rome. pp. 2659-2756.

205 Smith-Vaniz, W. F., 2002. Carangidae. In Carpenter, K. E. (Ed.). The living marine resources of the Western Central Atlantic. Vol. 3: Bony fishes part 2 (Opistognathidae to Molidae), sea turtles and marine mammals. FAO Species Identification Guide for Fishery Purposes and American Society of Ichthyologists and Herpetologists Special Publication No. 5. FAO, Rome. pp. 1426-1468.

Stevick, P. T., Aguayo, A., Allen, J., Avila, I. C., Capella, J., Castro, C., Chater, K., Dalla Rosa, L., Engel, M. H., Félix, F., Flórez-González, L., Freitas, A., Haase, B., Llano, M., Lodi, L., Munoz, E., Olavarría, C.Y., Secchi, E., Scheidat, M. and Siciliano, S., 2004. Migrations of individually identified humpback whales between the Antarctic Peninsula and South America. Journal of Cetacean Research and Management 6: 109-113.

Stewart, J., 2008. Capture depth related mortality of discarded snapper (Pagrus auratus) and implications for management. Fisheries Research 90: 298-295.

Stewart, J., Ferrell, D. J. and van der Walt, B., 2004. Sizes and ages in commercial landings with estimates of growth, mortality, and yield per recruit of yellowtail kingfish (Seriola lalandi) from New South Wales, Australia. Marine and Freshwater Research 55: 489-497.

St. John, J., 2003. Is your fish ‘bent’ and will it survive? SPC (Secretariat of the Pacific Community), Live Reef Fish. Information Bulletin 11: 31-35.

St John, J. and Syers, C. J., 2005. Mortality of the demersal West Australian dhufish, Glaucosoma hebraicum (Richardson 1845) following catch and release: The influence of capture depth, venting and hook type. Fisheries Research 76: 106- 116.

St John, J., King, J. and Saville, K., 2005. West Coast Demersal Scalefish Fishery. status report. In: Fletcher, W. J. and Head, F. (Eds.), State of the Fisheries Report 2004/ 05. Department of Fisheries, Western Australia.

Strand E., Jorgensen C. and Huse G., 2005. Modelling buoyancy regulation in fishes with swimbladders: bioenergetics and behaviour. Ecological Modelling 185: 309-327

Sumner, N. R. and Williamson, P. C., 1999. A 12-month survey of coastal recreational boat fishing between Augusta and Kalbarri on the west coast of Western Australia during 1996-97. Fisheries Research Report No 117. Fisheries Western Australia, Perth.

Taylor, R. G. and White, J. A., 1992. A meta-analysis of hooking mortality of nonanadromous trout. North America Journal of Fisheries Management 12: 769- 767.

Thompson, B. A., Beasley, M., and Wilson, C. A., 1999. Age distribution and growth of greater amberjack, Seriola dumerili, from the north-central Gulf of Mexico. Fishery Bulletin 97: 362-371.

Thresher, R. G., 1984. Reproduction in reef fishes. T.F.H. Publications Pty Ltd, Neptune City, New Jersey.

206 Thunberg, E. 1999. Marine recreational fishing participation in the northeast, U.S. 2000- 2025. In: Pitcher, T.J. (Ed) Evaluating the Benefits of Recreational Fishing. The Fisheries Centre Research Report 7 (2), The Fisheries Centre, University of British Columbia, Vancouver. pp.143-149.

Tserpes, G. and Tsimenides, N., 1995. Determination of age and growth of swordfish Xiphias gladius L., 1758, in the eastern Mediterranean using anal-fin spines. Fishery Bulletin 93: 594–602.

Uehara, S., Taggart, C. T., Mitani, T. and Suthers, I. M., 2006. The abundance of juvenile yellowtail (Seriola quinqueradiata) near the Kuroshio Current: the roles of drifting seaweed and regional hydrography. Fisheries Oceanography 15: 351-362. van der Walt, B., Faragher, R. and Lowry, M., 2005. Hooking mortality of released Silver Perch (Bidyanus bidyanus) after capture by hook-and-line fishing in New South Wales, Australia. Asian Fisheries Science 18: 205-216.

Wang, Y. X., Heigenhauser, G. and Wood, C. M., 1994. Integrated responses to exhaustive exercise and recovery in rainbow trout white muscle: acid-base, phosphagen, carbohydrate, lipid, ammonia, fluid volume and electrolyte metabolism. Journal of Experimental Biology 195: 227-258.

Walsh, C., McKenzie, J., McGregor G., Poortenaar, C., Hartill, B. and Smith, M., 2003. Information available for the management of New Zealand kingfish (Seriola lalandi lalandi) stocks. New Zealand Fisheries Assessment Report 2003/25. 57 pp.

Ware, D. M. and Tanasichuck, R. W., 1989. Biological basis of maturation and wave spawning in ( harengus pallasi). Canadian Journal of Fisheries and Aquatic Science 46: 1776-1784.

Weihs, D., 1973. Hydromechanics of fish schooling. Nature 241: 290-291.

Weihs, D., 1975. Some hydrodynamical aspects of fish schooling. In: Wu, T.Y., Brokaw, C. J., and Brennen, C. (Eds) Symposium on swimming and flying in nature. Plenum Press, New York. pp. 703-718.

Wells, R. J. D. and Rooker, J. R., 2004. Distribution, age, and growth of young-of-the year greater amberjack, Seriola dumerili, associated with pelagic Sargassum. Fishery Bulletin 102: 545-554.

Wendelaar Bonga, S. E., 1997. The stress response in fish. Physiological Reviews 77: 591- 625.

West, G., 1990. Methods for assessing ovarian development in fishes: a review. Australian Journal of Marine and Freshwater Research 41: 199-222.

Wilbur, H. M. and Rudolf, H. W., 2006. Lige-history evolution in uncertain environments: bet hedging in time. American Naturalist 168: 398-411.

Wilkie, M. P., Brobbel, M. A., Davidson, K., Forsyth, L. and Tufts, B. L., 1997. Influences of temperature upon the postexercise physiology of Atlantic salmon (Salmo salar). Canadian Journal of Fisheries and Aquatic Science 54: 503-511.

207

Wilson, R. R. and Burns, K. M., 1996. Potential survival of released groupers caught deeper than 40 m based on shipboard and in-situ observations, and tag-recapture data. Bulletin of Marine Science 58: 234-247.

Winstanley, R. H., 1990. A fisheries manager’s application of minimal legal lengths. In: Hancock, D. A. (Ed), Minimum legal sizes and their use in management of Australian fisheries. Australian Society for Fish Biology Workshop Proceedings. Australian Government Publishing Service, Canberra. pp. 51-56.

Wiseman B., 2003. Yellowtail kingfish - an emerging industry. Western Fisheries, Winter 2003: 46-48.

Wittenberg, B., Schwend, M. J. and. Wittenberg B. A., 1964. The secretion of oxygen into the swim-bladder of fish. III. The role of carbon dioxide. Journal of Physiology 48: 337-355.

Zuyev, G. V. and Belyayev, V. V., 1970. An experimental study of the swimming of fish in groups as exemplified by the horsemackerel (Trachurus mediterraneus ponticus Aleev). Journal of Icthyology 10: 545-549.

208 Appendix

Datasheet used by recreational anglers during the Samson Science project for the tag and release of Seriola hippos.

_ _ :

:: :: :

Start TimeStart Time Finish eg.bleeding, details of old tag, tag, old of details eg.bleeding, 2 3 1 ts thod, signs of barotrauma etc barotrauma of signs thod, e thod, hook position, details of release release of details position, hook thod, ______mor

Commen me lift other me revive or

or10 mins Fishing Location Fishing Location Fishing Location change if move of if move change ______Release _ _ Condition Note: Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Healthy Floated/Revived Died Crew Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Towed hose Deck Release wt Speared Speared Towed hose Deck Release wt Towed hose Deck Release wt Revive Revive Method Lift Lift _ Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Tail Other Net Leader Net Leader Tail Other Tail Other Method m _ Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Mouth Gut Other Hook Postn Line ______less 15kg less 15-30kg 30+kg less 15kg less 15-30kg 30+kg less 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg 15kg less 15-30kg 30+kg _ Jig Jig Jig Jig Jig Jig Jig Jig Jig Jig Jig Jig Jig Jig Jig Bait Bait Bait Bait Bait Bait Bait Bait Bait Bait Bait Bait Bait Bait Bait Fishing Method Class ______Fork (in cm) (in Length Skipper Fish Lift Height _ If required If _ c/o WAFMRL, PO Box 20, NORTH BEACH WA 6920 WA Box 20, NORTH PO BEACH c/o WAFMRL, Tag 1 Tag 2 Mob: 0418 326 747(Drew), 0427 774 551(Paul), 0427 472 121 (Mike) 0427 774 551(Paul), 747(Drew), 3260418Mob: Samson Science Tagging Datasheet Samson Science Tagging ______Tagger Number Number Please return sheets or direct queries to Andrew Rowland, Paul Lewis or Mike Mackie Mike or Lewis Paul Rowland, to Andrew queries direct or sheets return Please (If not main) not (If ______essel V Tagger Main _ _ (m) Depth Angler ______/ ______/ _ and Latitude ___ Longitude _ Date Tag Colour 1 2 3 4 5 6 7 8 9 14 15 10 11 12 13

209